Primary care physicians who are academically inexperienced are often recruited to fill part- and full-time faculty positions at medical schools and affiliated teaching sites. Since the 1970s, faculty development programs (FDPs) have been designed and implemented to promote the academic skills of new and junior primary care physicians and socialize them to the cultures of academic medicine.1 One key pathway to faculty development and academic medicine socialization is the development of a network of professional academic colleagues.2–4
Academic colleagues have long been associated with career benefits for higher education faculty. Since 1990, research has shown positive associations between the presence of academic colleagues and primary care faculty benefits, including gains in knowledge of the unspoken values and expectations of academic medicine, greater job satisfaction, and retention in academic roles.2,3,5,6 Those who often consult their colleagues about research and other academic projects generate more publications and other scholarly products.3,7 In fact, a 1995 literature review of colleague networks found that the evidence of their importance “is so compelling that if one were allowed only one line of inquiry to predict a faculty member's future success in the field, it might well be ‘Tell me about your colleagues.’”3
Recently, FDP curricula have been designed to encourage the important role colleagues can play in academic socialization and career development, and several single-program studies have reported changes in colleague network composition as an outcome associated with FDP participation.5,6,8 While these studies show overall positive associations between FDP participation and the development of colleague relationships (CRs), several concerns limit their impact. First, these studies have focused on single programs and small sample sizes; therefore, it is not clear whether enrollees across many FDPs have experienced colleague network gains. Second, the literature on colleague networks indicates that three types of academic colleagues form one's professional network: mentors, peers, and academic consultants.3,9,10 However, previous studies have focused almost exclusively on one type of colleagues, mentors, which ignores potentially important network members such as peers and consultants, and unrealistically limits understanding and opportunities to enhance faculty development practice in this area. Finally, prior studies have not reported specific benefits for FDP participants associated with CRs.
We therefore designed this nationwide study of family medicine FDP participants to answer two main research questions: (1) How many and what types of CRs emerge from FDP enrollment? and (2) What career benefits are associated with CR totals that emerge from FDPs?
We conducted a retrospective, cross-sectional study in two phases. During phase one we gathered detailed information from FDP directors about program features and enrollee rosters needed for phase two. In phase two, enrollees reported the numbers and types of CRs that emerged from FDP participation and the career benefits associated with CR totals. We based written questionnaires for both phases on Dillman's total design method that details a system of techniques for maximizing survey quality and returns.11
Participants in the program director survey were all family medicine FDP directors (n = 52) who received support from the U.S. Department of Health and Human Services, Health Resources and Services Administration (HRSA), during one or more years between July 1994 and June 1997. HRSA-funded programs were selected because these awards demonstrated the programs had met established standards for quality and rigor (e.g., clearly articulated objectives, an evaluation plan, procedures for following up with graduates).12
The FDP director survey underwent two pretests with purposive samples of FDP directors selected for their expertise and because they represented a range of FDP types (e.g., one-year, community-based program, two-year, full-time residency focus). We administered the questionnaire between July and September 1998 and asked the directors to provide numerical answers (e.g., “What was the planned, overall length of enrollment in this FDP?”). We also asked them to provide the number of annual enrollees and the number of years the program had operated. We included instructions for preparing and returning complete rosters of qualified enrollees—family physicians who had participated in HRSA-funded FDP activities for six months or longer during the 1994–1997 time frame. These qualified enrollees listed on the director-provided rosters were the subject of phase two.
Phase two concerned the enrollee questionnaire, which was pretested with a non-overlapping subset of primary care FDP graduates prior to its being administered to the participants. We mailed up to three survey packets to 616 qualified enrollees, the first in late October 1998. We sent a postcard-type questionnaire to non-respondents in early February 1999 and collected all respondent and non-respondent data by the end of February 1999.
The enrollee questionnaire consisted of 51 items divided into three sections. In the first section participants used a five-point Likert scale to rate the degrees to which their FDPs had helped them improve key socialization skills taken from a model curriculum for physician faculty development.2 In the second section participants reviewed the descriptions of three CR types adapted from academic and corporate career literature,9,10 and entered the total number for each type with whom they had “initiated or strengthened a professionally helpful relationship directly because of FDP participation.” The CR types were:
* Mentors, who are often in advanced career stages. They foster academic identity, confidence, and career advancement. Mentors might link someone to new opportunities and people, or advise on career goals.
* Peers, who are often at similar career stages. They provide informal feedback and friendship. With peers a faculty member might discuss difficult students or team teach a continuing medical education workshop.
* Academic consultants, who provide specialized help in activities and projects that aid efficiency and quality. Academic consultants might assist with data analysis or edit meeting abstracts.
In addition to the above CR types that had already provided career support, participants also entered a total number for FDP colleagues who had not yet provided career help but were perceived likely to do so “if asked in the coming year.” We included these perceived colleagues because of the psychosocial benefits attributed to them.13 Participants were instructed to enter each FDP colleague only once in the above categories.
The third section of the questionnaire required participants to identify specific details about the one or two most helpful CRs from the above categories. We asked participants whether and how these most helpful CRs had supported one or more achievements (e.g., teaching achievements, publications, grant preparation) linked with faculty advancement.2,10 Participants also reported whether these CRs were primarily affiliated with external institutions, and the strengths of their connections to regional or national academic or research networks. These affiliations and connections have been associated with ongoing career support and achievement.3,13
We used analysis of variance (ANOVA) to compare participating and non-participating FDP programs on duration, size, and years of operation and checked the instrument's test—retest reliability for the enrollees using Pearson's correlation. Factor analysis helped ensure the validity of the socialization scale.14,15 We analyzed the two main research questions with descriptive statistics, chi-square, and ANOVA.14,15 We used the Tukey post-hoc test for multiple comparisons, and a significance level of .05 as a criterion for statistical significance.15
Of the 52 directors contacted, 49 returned completed instruments with information about their program characteristics. Twelve of those did not return roster information, most reporting that their institutions prohibited releasing faculty names for research purposes. Therefore, 37 of the 52 directors (71%) returned completed questionnaires and enrollee rosters. Participating and non-participating programs were similar on each of three comparison variables (years in operation, program duration, and annual enrollees). The 37 participating programs were dispersed throughout the ten U.S. regions that the HRSA uses to geographically categorize funded programs.
Findings for Enrollees, Instrument Reliability, and Validity
Non-deliverable mail and returns from those who were ineligible (e.g., non—family physicians) resulted in a final population frame of 543 participants. A total of 351 useable instruments were returned, a 65% response rate. A comparison of respondents who returned the study questionnaire and those who returned a postcard follow-up only (n = 59) showed no significant difference in gender, racial or ethnic background, or program duration (p > .05).
The final study cohort comprised 229 men (65%) and 122 women (35%), whose mean age when starting an FDP was 39 (SD = 7.1). The participants' average number of years as faculty members before starting an FDP was 3.8 (SD = 5.4), with the women having an average of 2.7 years of faculty experience compared with the men's average of 4.4 years. The respondents' racial or ethnic backgrounds were 84% white, 5% Hispanic, 4% Black, 4% Asian, and 3% “other.” When comparing the majority (white) with the minority (all non-white ethnic or racial backgrounds), minority respondents averaged 1.9 years as faculty prior to beginning their FDPs, while majority faculty averaged 4.2 years.
Test—retest analysis showed the overall correlation of enrollee survey questions was .86, exceeding the target correlation of .70.14,15 Test—retest analysis of item subsets found that items focused on differentiating colleague types had a correlation of .58. Reanalysis of this subset of responses using total CRs (collapsing across types) improved the correlation of this subset of items to .86 (for more detailed analysis, see Discussion).
Factor analysis resulted in a three-item scale for socialization skills, consistent with the model of academic physician socialization on which it was based.2 Scale items were the ability to manage a productive career in academia; the ability to understand the norms, values, and unwritten rules of academia; and skills to build and maintain a network of academic colleagues. The loadings (between .59 and .79) for these factors indicate they strongly affected the socialization construct.14 Extraction sums of squared loadings showed these factors explained 52.4% of the construct's variance. The Chronbach's alpha of .76 indicated the acceptability of the scale's internal consistency.14,15
Findings on Research Questions
Research question one asked “How many and what types of CRs emerge from FDP enrollment?” The participants had developed a mean of 9.1 academic CRs associated with FDP enrollment. Just over six of these colleagues had provided actual career support (on average, three were categorized as peers, two as mentors, and 1.4 as academic consultants). Three additional colleagues were perceived “likely to provide career support” if asked in the coming year. No statistically significant difference existed in colleague types or totals by racial or ethnic background or gender, but some differences in CR totals were noted. Asian and Black participants reported 7.8 and 8 CR totals, which were fewer than those of white and Hispanic participants, who reported 9.2 and 9.8 CR totals, respectively. Women and men reported almost identical CR totals (9.2 and 9.1).
Research question two asked “What career benefits are associated with CR totals?” A positive and significant association existed between CR totals and socialization improvement ratings (see Table 1). However, post-hoc analysis indicated socialization increases were not significantly different when participants reporting ten through 17 CRs were compared with those reporting six through nine or over 17.
We also asked participants to compare their FDP colleagues with their overall colleague networks and make a relative judgment of the FDP colleagues' “importance to academic career.” A total of 69% of participants stated their FDP colleagues had been important or very important, 26% were unsure, and 5% said they had been unimportant when compared with their overall academic networks.
Additional benefits from CRs related to the participants' academic achievements and products. They described 527 “most helpful” CRs that had contributed to 898 products or achievements (e.g., improved administration, regional and national presentations) (see Table 2). Forty-three percent of these most helpful colleagues were primarily affiliated with external institutions, and 74% were very well or well connected to regional or national networks of scholars.
These FDP enrollees had initiated or strengthened over nine CRs due to FDP participation. Among the colleagues providing support, participants reported peers most frequently, followed by mentors, then academic consultants. While the career development literature appropriately focuses on mentors, our findings support additional study into the characteristics and functions of peers and academic consultants. In addition, these findings support continuing research emphasis on the composition and influence of developmental clusters of diverse CRs rather than on single developmental relationships.16 Results in the fourth category of colleagues studied, perceived CRs, suggest FDP participants believe support is available from three additional contacts formed in FDPs. Perceived CRs are positive additions to one's CR network because they are associated with job satisfaction without the costs of reciprocating that are associated with exchange-based work relationships.13
The lack of significant differences between the CR totals of men and women as well as between minority and majority participants is encouraging but should be interpreted with caution. Quantitative research methods such as those used in this study may not reveal important CR differences existing within and across these subgroups.
Consistent with studies based on higher education and business, as CR totals increased, so did the mean scores for socialization improvement.3,9,10 Because academic socialization is crucial for faculty success,1,3 this finding empirically supports the continued focus on colleagues as a vehicle for professional socialization. However, we found evidence of a plateau effect of CR totals on socialization gains, as two post hoc comparisons of upper-level gains did not reach statistical significance. Beyond CR totals of about ten, there are diminishing socialization returns from CRs initiated or strengthened in FDPs. Prior researchers have also expressed concern about increasingly large support networks that can require increasing obligations to extend career help to those who have provided it.13
This study found the participants' most helpful CRs had each aided almost two academic products or achievements. This important finding should not be over-interpreted, as the most helpful one or two CRs may not represent the support provisions of one's entire network. Still, we found the measurable contributions of these CRs important to the enrollees' academic projects and achievements. Beyond their help on projects, these most helpful CRs may significantly influence the enrollees' future careers due to the study's findings relative to the prevalence of their affiliations with external institutions and links to influential networks.
This study has limitations. It presents participants' retrospective views of program impact, which may threaten the study's validity. The study used self-report data, which can be a biased indicator of experiences. However, the anonymity of self-completed surveys may make them effective in gathering data on sensitive topics such as relationship development.14 Finally, the responses of test—retest participants on the instrument section “colleague relationship types” did not reach the target reliability. However, prior research shows CRs naturally evolve in their influences and types (e.g., from mentor to peer).9 We believe that this view of evolving CR types is supported by the test—retest reliability of total CRs.
In summary, the results of this study extend prior research on FDPs and CRs. FDPs are positively associated with initiating and strengthening relationships with mentors, peers, and academic consultants who provide actual career support and are perceived to be available for future career assistance. Important career benefits, such as skills to manage one's career and academic achievements and products that foster career advancement, are linked to these CR gains.
1. Bland CJ. Guidelines for planning faculty development workshops. J Fam Pract. 1977;5:235–41.
2. Bland CJ, Schmitz CC, Stritter FT, Henry RC, Aluise JJ. Successful Faculty in Academic Medicine. New York: Springer-Verlag, 1990.
3. Hitchcock MA, Bland CJ, Hekelman FP, Blumenthal M. Professional networks: the influence of colleagues on the academic success of faculty. Acad Med. 1995;70:1108–16.
4. Baldwin CD, Levine HG, McCormick DP. Meeting the faculty development needs of generalist physicians in academia. Acad Med. 1995;70(1 suppl):S97–S103.
5. Morzinski JA, Diehr S, Bower DJ, Simpson DE. A descriptive, cross-sectional study of formal mentoring for faculty. Fam Med. 1996;28:434–8.
6. Woods SE, Reid A, Arndt JE, Curtis P, Stritter FT. Collegial networking and faculty vitality. Fam Med. 1997;29:45–9.
7. Jones JE, Preusz GC, Finkelstein SN. Factors associated with clinical dental faculty research productivity. J Dent Educ. 1989;43:638–45.
8. Marks MB. Academic careers in medical education: perceptions of the effects of a faculty development program. Acad Med. 1999;74(10 suppl):S72–S74.
9. Kram KE. Mentoring at Work: Developmental Relationships in Organizational Life. Glenview, IL: Scott Foresman, 1985.
10. Hill SK, Bahniuk MH, Dobos J, Rouner D. Mentoring and other communication support in the academic setting. Group and Organizational Studies. 1989;14:355–68.
11. Dillman DA. Mail and Telephone Surveys. New York: John Wiley & Sons, 1978.
12. Health Resources and Services Administration. Bureau of Health Professions Uniform Progress Report. OMB No. 9015-0061. Rockville, MD, 1999.
13. Vaux A. Social Support: Theory, Research, and Intervention. New York: Praeger, 1988.
14. Survey Research Using SPSS, Statistical Package for the Social Sciences. Chicago, IL: SPSS, 1998.
15. Hays WL. Statistics. 4th ed. Fort Worth, TX: Holt, Rinehart & Winston, 1988.
16. Higgins MC, Kram KE. Reconceptualizing mentoring at work: a developmental network perspective. Academy of Management Review. 2001;26:264–88.