Training and supporting the pool of mental health researchers required for a thriving research community is difficult in part because of the length of time between education and research career independence (1). For example, the average age for receiving a National Institutes of Health independent research grant was 36 in 1981, but 42 in 2010 (2). During this time, financial and professional pressures can drive talented individuals out of research and into other fields (3).

Career-development conferences contribute to the retention of high-quality researchers by providing information specific to a research career, such as the grant submission process, as well as information generally applicable to all early-career professionals, such as finding a work–life balance (4). Conferences can also offer “survival skills” training by providing techniques for and examples of successful research career skills such as writing and negotiation strategies (5). Also, programs can facilitate the development of mentoring skills and relationships in a context where both senior and junior scholars can benefit (6, 7).

Although career-development conferences provide an established forum for continuing education, the costs of attending conferences in person can be prohibitive (8–10). Online conference dissemination may be a viable method for extending the value of career-development conferences. Conference websites, for example, can reduce registration and travel costs and increase international exposure for conference materials and ideas (11). Online conference material also allows researchers to fit educational time into a busy schedule and supports different learning preferences (12).

Previous work has shown that conference materials adapted for online training can yield similar knowledge gains when compared with in-person training (13). These findings are consistent with the “no significant difference” theory, which holds that qualities other than the training medium drive success for online education (14). In other words, online and in-person instruction are generally both effective for conveying knowledge when the same instruction methods are used (see references 15–18 for reviews).

Three objectively measurable research skills are well recognized as indicators of scientific development: running successful research projects, writing publications, and securing honors or grant awards. Several of the major career-development institutes explicitly list writing and earning grants in their mission statements (19–21). Furthermore, grants completed, papers published, and honors received are all used to measure excellence of scientific contributions during tenure decisions (22–24).

Scientific collaborations and presentations also provide beneficial experience to early-career researchers. Collaborations can be formal or informal and range from brainstorming to results- or expertise-sharing (25). Sharing data and experiences is particularly useful because it can influence scientific perspectives and challenge assumptions, especially at international levels (26). Collaboration may also improve researchers' chances for publication (27), a critical yet often challenging accomplishment for early-career researchers. Finally, scientific presentations emphasize both the cognitive and the social elements of career development (28) by offering professional and networking benefits for the individual researcher as well as providing a forum for disseminating scientific findings.

The current study examines online and in-person career-development training and the career advancement of attendees. Data are presented from 106 trainees who participated in three different career-development institutes: the 2008 Career Development Institute (CDI) for Bipolar Disorder, the 2008 CDI for Psychiatry, and the 2008 Leadership Training Institute (LTI). The work contributes to the question of how online and in-person research training can positively affect career change.

Participants

Participants in the current study included 106 trainees who participated in one of three career-development institutes, each of which took place during the spring of 2008: CDI for Bipolar Disorder (N=41), CDI for Psychiatry (N=41), and LTI (N=24). All participants were either early-career or mid-career researchers. Of the 106 participants, 75% percent of the participants were early-career, and 25% were mid-career. “Early-career” was defined as having obtained a terminal degree, having fewer than 10 peer-reviewed publications, and either having or seeking full-time employment in a research career (e.g., applying for a K award; first-year assistant professor status). Mid-career was defined as having obtained a terminal degree, having a broader range and higher number of peer-reviewed publications than early-career individuals, possibly having received some research funding but not yet being fully independent, and having some experience in a research career (e.g., receiving an R03 award; associate professor status).

The sample of trainees included 66% women and 34% men with a self-identified racial/ethnic distribution of 71% White, 15% Asian, and 14% Black, with 14% of the total sample reporting being of Hispanic descent. Chi-square analyses showed no significant differences across the three conferences by gender, career level, or ethnicity. However, significant differences in race were found (χ²[4]=35.30; p <0.0001) where more Black (50% for the LTI versus 4% for the CDIs) and fewer White trainees (33% for the LTI versus 81% for the CDIs) participated in the LTI than in the other two conferences. This difference reflects the fact that the LTI was designed specifically for women and minority researchers.

Procedures

This study received expedited Institutional Review Board approval. Conference organizers solicited applications for their conference via national researcher list-servs, e-mail notifications to academic departments in relevant fields across the country, and advertisements in relevant professional organization publications. Each application required completion of an online form, submission of a CV, a statement of research experience and/or interests, and at least one letter of recommendation. More than 285 researchers applied for the three training programs, and attendees were competitively selected by the conference organizers, with a selection rate of approximately 50% of applicants.

Participants for this study were recruited from the list of conference attendees, as well as recommendations of conference faculty and direct e-mails to early- and mid-career researchers whose publications matched the conferences' respective areas of specialty. Study participants filled out an online or hardcopy consent form before completing pre-conference questionnaires. Participants had access to the career-development conferences through one of two conditions: 1) combination of in-person and online training (Combined: N=46), where participants attended the institute and also had subsequent access to the conference materials online; and 2) Web-Only training (Web-Only: N=60), where trainees had access to the conference materials-only online. Chi-square analysis showed no differences across the Combined and Web-Only conditions by conference, gender, race, Hispanic status, or career-level.

To maintain a high level of quality for the online component, live conference presentations were videotaped and adapted for online display (12). The online presentations provided viewing options that appealed to different learning preferences (e.g., full video presentation with transcript, slides and transcript only, transcript-alone). The mp3 audio files and slides for each presentation were also available for download. An online bulletin board, a downloads area, a links page, and faculty bio pages were also provided for each conference.

Measures

Participants completed a Career Progress Survey before and 9 months after participation in their respective conference. This timeframe represented approximately 1 year after acceptance into the study and completion of the pre-conference Career Progress Survey. This survey assessed different aspects of research career achievement and experiences, including current position, honors and awards, publications, career-development activities (e.g., seminars, training institutes), funding, and collaboration. Detailed information was collected to increase accuracy in reported information (e.g., publication citations, specific grant titles, and presentation title information).

Because of time and scheduling issues, 24 trainees dropped out over the course of the study (23% attrition rate). Chi-square analyses showed no evidence of selective attrition. Attrition analysis included initial level of career achievements, conference, study condition, gender, race, Hispanic status, or career level, and there were no significant differences between trainees who exited the study versus those who remained for any of these categories.

In the first set of analyses, the five key markers of career progress were examined: 1) number of publications (including peer-reviewed journal articles, book chapters, trade magazine articles, and abstracts); 2) number of grants on which the participant served as PI or Co-PI; 3) number of honors or awards; 4) number of current research collaborators; and 5) number of presentations. Each variable was collected at pre-assessment and again at a 9-month follow-up. A multivariate analysis of variance (MANOVA) across the five variables was run, with study Condition (Combined versus Web-Only) as the between-subjects variable and Time (pre-conference versus 9 months later) as the within-subject repeated factor. A main effect for Conference was significant at the multivariate level (F[20, 134]=2.13; p <0.05); therefore, Conference was included as a covariate and controlled for in all subsequent analyses.

Significant multivariate main effects for study Condition were found both at the pre-conference timepoint (F[5, 98]₎=2.91; p <0.05) and at 9-month follow-up (F[5, 71]=2.67; p <0.05; see Table 1). Univariate analyses revealed that participants in the Combined condition showed a significantly larger number of honors at pre-conference than did Web-Only participants (F[1, 102]=9.94; p <0.01), and participants in the Web-Only condition showed a trend toward higher number of collaborators at pre-conference than did the Combined participants (F[1, 102]=2.93; p<0.1). However, at 9-month follow-up, only number of honors continued to show this difference across study conditions (F[1, 75]=10.26; p <0.1). There were no significant multivariate interaction effects at either timepoint as a function of Conference, study Condition, Gender, Race, Hispanic status, or Career level.

Examination of the repeated-measures MANOVA indicated a significant multivariate effect for change over time from pre-conference to 9 months (F[9, 675]=18.00; p <0.0001). The Time effect did not vary by Condition. Univariate analyses revealed that this effect held for four of the five career-progress variables: number of publications (F[1, 75]=9.36; p <0.01), number of grants (F[1, 75]=6.95; p <0.01), number of research collaborations (F[1, 75]=3.95; p <0.05), and number of scientific presentations (F[1, 75]=11.11; p <0.01). Each marker of career progress significantly increased over time. Only the number of honors received did not show significant change over time. Also, there was a significant multivariate interaction effect for change over time by career level (F[9, 657]=1.90; p <0.05). The change in number of research collaborations varied for early-career versus mid-career level researchers (F[1, 73]=9.80; p <0.01). Specifically, least-squares means showed greater improvement for early-career individuals, from a total of 1.15 (SD: 0.67) at pre-conference to 4.15 (SD: 0.74) at 9 months, whereas mid-career researchers moved from a total of 3.92 (SD: 1.13) at pre-conference to 4.35 (SD: 1.26) at 9 months.

In a second analysis, a composite variable representing Total Career Progress was created based on demonstrated change in each career progress area. For each participant, a Total Career Progress score was calculated by allotting 1 point for each key marker that was greater at 9 months than it was at pre-conference, and an additional point was given for having a higher salary or a new job at follow-up. Descriptive analyses revealed that 57% of participants had an increase in salary, and 15% of Combined and 13% of Web-Only trainees had obtained a new position by the end of the study.

An analysis of variance (ANOVA) with study Condition as the between-subjects factor indicated no significant difference in Total Career Progress by Condition (Combined M=1.98; SD: 0.26; Web-Only M=2.03; SD: 0.23). However, a significant difference by Gender was found (F[1, 100]=8.12; p <0.01) where women showed greater overall career progress over time than did men (women: M=2.37, SD: 0.21; men: M=1.33, SD: 0.30). No other main or interaction effects were observed.

This study examined the career achievement of trainees before and 9 months after participation in a career-development conference. Trainees in social, behavioral, and mental health fields of research showed significant improvement in key career progress markers, with the number of presentations and publications showing the greatest degree of expansion. For Total Career Progress, female researchers showed the greatest overall career growth. Although the number of women graduating with terminal degrees has increased over time, a commensurate increase in leadership and faculty positions held by women has not yet been seen (29). The findings suggest that women with access to a career-development conference may show more proactive career behaviors than their male counterparts and thereby be in a stronger position for career advancement. This study supports the promise of career-development programs, such as the CDIs and LTI, for supporting the career trajectory of female researchers in particular.

Similar to previous studies (15–17), no significant differences in career progress over time were evident across the two study conditions, as measured through either individual markers or the Total Career Progress index. Thus, the combined in-person-plus-web-training group and the web-only training group each showed significant career progress regardless of how trainees participated in the conferences. This finding is congruent with previous work showing “no significant difference” in the benefits of online versus in-person training for knowledge acquisition (14) and extends this argument to suggest equal benefits with regard to career-enhancing skills and experiences. Although more work is needed regarding the degree to which career-development conferences directly contribute to improvements in these key career progress markers, this study suggests that online conferences offer equally potent career-enhancing opportunities for researchers. As has been found in past research comparing different training modalities, benefit for trainees is most likely determined by the quality of the underlying training materials, and not the particular training modality (14, 15).

Limitations of this study include its small sample size and the lack of a control condition that did not receive any training. Thus, the degree to which trainee career-development can be attributed to conference participation is limited. Questions remain about the ways in which time spent in various conference participation methods and individual initiative help produce successful career practices. Future research is needed with a larger set of trainees in order to explore how these factors interact with one another to determine career progress for researchers over time. Also, this study’s timeframe was limited to approximately 1 year between pre-conference and follow-up assessment, which is a relatively short period in which to see substantial career growth. Whereas significant changes in key markers of career success were evident in this study, it is important that future research examine the longer-term impact of career-development conferences for career achievement over multiple years. In particular, given that online conferences offer fewer, or at least different, opportunities to obtain and engage with mentors, future studies should examine whether the impact of online educational programs over a longer career span differs from that of in-person programs.

In this difficult funding climate, when training programs face increased demands for justification, the measure of career advancement is particularly significant. Although a conference cannot be the sole form of support for trainees, this study bolsters the argument that career-development institutes can support significant positive change in career progress markers for trainees. In particular, online career-development conferences can provide affordable, feasible training opportunities to support the retention and career-enhancement of researchers on a broad scale. Given that online training modalities offer lower participation barriers and greater cost-effectiveness than in-person methods, this study underscores the potential value of offering high-quality, web-based career-development conferences as a viable means to increase access to career-development training opportunities for researchers.

The authors thank the faculty and trainees of the Career Development Institutes and the Leadership Training Institute, without whom none of this research would have been possible. In particular, Dr. David Kupfer's guidance and commitment to research training are greatly appreciated.

This research was funded in part by the National Institute of Mental Health (contracts HHSN271200774104C, HHSN278200553104C, and HHSN271200774100C).