Transformational, Transactional, and Passive-Avoidant Leadership Characteristics of a Surgical Resident Cohort: Analysis Using the Multifactor Leadership Questionnaire and Implications for Improving Surgical Education Curriculums

Journal of Surgical Research 148, 49 –59 (2008) doi:10.1016/j.jss.2008.03.007

Transformational, Transactional, and Passive-Avoidant Leadership Characteristics of a Surgical Resident Cohort: Analysis Using the Multifactor Leadership Questionnaire and Implications for Improving Surgical Education Curriculums Irwin B. Horwitz, Ph.D.,*,1 Sujin K. Horwitz, Ph.D.,† Pallavi Daram, MPH,* Mary L. Brandt, M.D.,‡ F. Charles Brunicardi, M.D.,‡ and Samir S. Awad, M.D.‡ *Department of Management Policy and Community Health, University of Texas School of Public Health, Houston, Texas; †Department of Management and Marketing, University of St. Thomas, Cameron School of Business, Houston, Texas; and ‡Michael E. DeBakey Department of Surgery, Baylor College of Medicine, Houston, Texas Submitted January 6, 2008

Leadership Questionnaire was demonstrated to be a valuable tool for identifying specific areas where leadership training would be most beneficial in the educational curriculum. The future use of this instrument could prove valuable to surgical education training programs. © 2008 Elsevier Inc. All rights reserved. Key Words: surgical education; leadership; medical curriculum, training; communication; ACGME core competency.

Introduction. The need for leadership training has become recognized as being highly important to improving medical care, and should be included in surgical resident education curriculums. Materials and methods. Surgical residents (n ⴝ 65) completed the 5x-short version of the Multifactor Leadership Questionnaire as a means of identifying leadership areas most in need of training among medical residents. The leadership styles of the residents were measured on 12 leadership scales. Comparisons between gender and postgraduate year (PGY) and comparisons to national norms were conducted. Results. Of 12 leadership scales, the residents as a whole had significantly higher management by exception active and passive scores than those of the national norm (t ⴝ 6.6, P < 0.01, t ⴝ 2.8, P < 0.01, respectively), and significantly lower individualized consideration scores than the norm (t ⴝ 2.7, P < 0.01). Only one score, management by exception active was statistically different and higher among males than females (t ⴝ 2.12, P < 0.05). PGY3-5 had significantly lower laissez-faire scores than PGY1-2 (t ⴝ 2.20, P < 0.05). Principal component analysis revealed two leadership factors with eigenvalues over 1.0. Hierarchical regression found evidence of an augmentation effect for transformational leadership. Conclusion. Areas of resident leadership strengths and weaknesses were identified. The Multifactor

INTRODUCTION

The role of surgeons has transcended that of exclusively acting to conduct surgical procedures, and now includes the necessity of incorporating the qualities of effective communication skills, managerial professionalism, and emotional regulation in providing comprehensive medical care. Together, such abilities under the rubric of “leadership” have become so important that the Accreditation Council for Graduate Medical Education (ACGME) now considers several elements of leadership as core competencies required as part of medical training. For example, among surgeons the need for leadership skills has become significant enough to have been recognized as the “new mandate” in surgical training [1], yet so underdeveloped that it has also been argued to be the Achilles’ heel among academic surgeons [2]. Throughout the medical literature, the concurrence favoring physician leadership training among medical scholars is widespread [3–11]. Leadership training among medical residents has been found to lead to improvements in overall health care,

1 To whom correspondence and reprint requests should be addressed at Department of Management, Policy, and Community Health, University of Texas School of Public Health, 1200 Herman Pressler, E303, Houston, TX 77030. E-mail: Irwin.Horwitz@uth. tmc.edu.

49

0022-4804/08 $34.00 © 2008 Elsevier Inc. All rights reserved.

50

JOURNAL OF SURGICAL RESEARCH: VOL. 148, NO. 1, JULY 2008

including enhanced communication [12, 13], facilitating team building goals and coordination among medical teams [14 –18], medical error reduction [19 –24], and gaining patient trust, compliance, and satisfaction [25–28]. However, irrespective of the many advantages to including leadership programs in academic curriculums and health care institutions, several barriers act to negate the internalization of leadership principles or their integration into practice. These include traditional tension between physicians and nurses [29, 30], ambiguous goal alignment and low morale [31], ineffectiveness of convincing residents of the importance of leadership has to their clinical practice [2, 3], and difficulties inherent to attending physicians having to assume the dual role of teaching both medical and leadership skills to their residents [32]. For over the past 50 years, leadership as a managerial and academic subject of study has produced a wide number of theories [33– 40], the majority of which are based in trait, behavioral, or power-based taxonomic foundations. Other theories, which originally emerged in the assessment of follower maturity [41– 43], have progressed into theories of emotional intelligence [44 – 45], and has been preliminary examined with relevance to clinical skills [46]. Numerous critiques have pointed out the challenges of transference of emotional intelligence, the inability to measure this capability, and its independence of other constructs such as cognition and ability [47]. In fact, the definition of “leadership” itself is widely varied among organizational psychologists [48], though the most widely agreed elements of the general construct is that it involves a process of influence that an individual asserts over followers to attain specified goals [49, 50]. While many of the various models offer something of value for enhancing physician leadership, the most encompassing, given the specialization and variety of duties, are the intertwined theories of transactional and transformational leadership [51–53]. Because this model is primarily focused on the influence of follower beliefs and attitudes, it best reflects the intrinsic value-oriented philosophy of medical practitioners and, thus, incorporates patient centered values in the development of surgical resident leadership style. The related paradigms of transactional and transformational leadership have become among the most widely studied theories of leadership performance. Transactional leadership is the type of leadership a leader exercises when exchanging something of value to elicit a specified behavior from followers [54]. Transactional leadership is further subdivided into three areas: contingent-reward (CR), where rewards are provided given certain criteria are met; management by exception-active (MBEA) that aims to intervene with follower behavior before a course of action becomes problematic; and management by exception-passive

(MBEP) that does not interfere with follower behavior until a problem arises. Transformational leadership, in contrast, attempts to influence the beliefs and attitudes of followers to align with that of the leader, and then direct followers through these common beliefs toward the attainment of greater organizational success [54]. Four types of transformational leadership have been identified: (1) inspirational motivation (IM) aims to influence followers through charismatic communication of a set of goals that becomes viewed as universally valuable to achieve; (2) individualized consideration (IC) occurs when the leader serves to help the follower attain desired intrinsic needs; (3) idealized influence is commonly subdivided into two types: idealized influence attributed (IIA) in which leader charisma is used to foster strong positive emotional bonds with followers, and idealized influence behavior (IIB) in which the idealized behavior of the leader becomes manifested in collective values and actions throughout the organization; and (4) intellectual stimulation (IS) pushes followers to think creatively and pursue new and creative ideas. A less engaged approach, often separately deemed passive-avoidant leadership [55], is defined as laissez-faire (LF) leadership style, where leaders shy away from important decisions and abstain from an active leadership role. Additionally, the transactional variable of management by exception-passive is often considered to constitute passive-avoidant leadership under the same global construct as laissez-faire style leadership [55]. Table 1 provides a basic summary of the aforementioned leadership styles and basic components of which they are comprised. It is important to note that these two forms of leadership are not mutually exclusive. It is the composite relationship and balance between these nine elements that are held to determine the overall effectiveness of an individual’s leadership. One of the most important means by which such traits have come to be measured has been through the use of the Multifactor Leadership Questionnaire (MLQ) [55]. Originally consisting of 73 items, the MLQ was refined to a shorter and more precise version [56]. The MLQ version used in this study was the 5x-short, which is comprised of 45 items that measure rater beliefs about effective leadership behavior and outcomes. The MLQ is a validated instrument, and in the past has been used to assess a wide range of leadership behaviors among nursing populations [57], military personnel [58, 59], health care employees [60], and additionally examine health care quality outcomes given differences in general physician leadership orientations [61]. Moreover, in addition to the nine leadership factors, the MLQ also assesses raters’ perceptions of their effectiveness of their leadership style in three outcome domains. The first, effectiveness (EFF), is how effective the raters perceive themselves to be as leaders. The

HORWITZ ET AL.: LEADERSHIP CHARACTERISTICS OF A SURGICAL RESIDENT COHORT

51

TABLE 1 Summary of the Basic Leadership Styles and Their Components Leadership styles Transactional leadership

Characteristics

Descriptions

Contingent reward (CR) Management by exception-active (MBEA)

Transformational leadership

Idealized influence Attributed (IIA)

Idealized influence (behavior)

Inspirational motivation (IM) Intellectual stimulation (IS) Individualized consideration (IC) Passive-avoidant leadership

Management by exception-passive (MBEP)

Laissez-faire (LF)

Exchange-based leadership based on providing rewards and punishments based on follower behavior/performance Actively monitors situations and makes corrective interventions before situations become problematic (prospective) Provides vision and sense of mission, gains the respect and trust of followers, and subordinates become encouraged to emulate the behavior and values of the leader Sets example for follower behavior through own actions, subordinates look to behavior of the leader for guidance of their own behavior Utilizes referent power, communicates charismatically, clearly and unambiguously to followers Encourages creativity, thoughtfulness, rational action, and active problem solving among subordinates Provides individual attention, mentoring, empowerment, and bonding with followers Takes corrective action only after a problem becomes significant and obvious (retrospective). In some cases, can be considered a transformational leader quality when the leader purposely aims to let followers learn from making mistakes Decision avoidance, passes important decision making responsibility to subordinates, reluctance to express views on important or controversial issues

second, satisfaction (SAT), is a measure of how satisfied the raters are with their leadership outcomes. The third, extra effort (EE), is how much additional work the raters believe they are able to elicit from their subordinates with their leadership skills beyond stated organizational expectations. This third outcome can credibly be argued to be analogous to what has been described in the organizational psychology literature as organizational citizenship behavior [62]. This study examines data from the administration of the MLQ to a cohort of surgical residents. In doing so, it represents for the first time the composition of these leadership qualities and perceived outcomes have been conducted on a population of physicians undergoing surgical training. Through the analysis of these results, an initial basis for assessing the leadership skills of residents is explored, comparison to population averages contrasted, and particular areas in need of leadership education and training among surgical residents are identified.

demographic information page, which asked for information on the respondents age, gender, marital status, and postgraduate year (PGY). Respondents provided an assessment of their supervisory leadership behaviors by answering the 45 scale items on the MLQ. Each leadership behavior was rated on a 5-point Likert-type scale ranging from 0 (not at all) to 4 (frequently, if not always). Descriptive statistics were obtained to summarize characteristics of the surgical residents in aggregate and by demographic categorization. Bivariate analysis was performed to examine relationships between the variables in the study. To examine psychometric properties of the variables, reliability tests and a principal component analysis was also conducted. Finally, a series of univariate analyses and hierarchical regression analyses were performed to examine the relationships between the MLQ leadership dimensions and outcome variables. The data were also examined for extreme outliers and missing information. In the case where a missing subscale was detected, missing variable data were imputed for MLQ subscale items by using the mean score derived from subscale responses. When there were more than two missing subscale items, the entire subscale was considered as missing. All statistical analyses in this study were performed on SPSS version 14.0, 2005 (Chicago, IL).

MATERIALS AND METHODS

Demographic and Outcome Analysis

Surgical residents at Baylor College of Medicine’s Michael E. DeBakey Department of Surgery (n ⫽ 65) completed the MLQ 5xshort form (leader version). The residents were provided an informed consent form, told all participation was voluntary, and subject identification was not recorded. The study was approved by the Institutional Review Board of Baylor College of Medicine. Prior to the administration of the MLQ, the residents first filled out a general

Of the 65 surveys used, 43 (66.2%) were from males and 22 (33.8%) were from females. The average age of the resident respondents was 29.0 y (SD ⫽ 3.4). Of the resident population, 32 (49.2%) were PGY1, 7 (10.8%) were PGY2, 12 (18.5%) were PGY3, and PGY4 and PGY5 both were comprised of 7 (10.8%) residents each.

RESULTS

52

JOURNAL OF SURGICAL RESEARCH: VOL. 148, NO. 1, JULY 2008

A ␹ 2 test was conducted to examine if there was any significant difference by gender between the dichotomized PGY groups, and no significant difference was found (␹ 2 ⫽ 1.23, df ⫽ 1, P ⫽ 0.27). The 12 leadership scores, including the three outcome variables, were compared for statistically significant differences by gender and PGY. By gender, only one score, management by exception-active, was statistically higher at 2.46 for male residents and 1.96 for female residents (t ⫽ 2.12, df ⫽ 57, P ⬍ 0.05). Because PGY was so heavily skewed toward the first year, and residents in their third year and above are assumed to have acquired greater surgical mastery and given more responsibility, for the purposes of testing, PGY was dichotomized into PGY1-2, and PGY3-5. Statistical comparison of mean test scores revealed that the mean contingent reward score of 3.17 for those PGY 3–5 were significantly higher than the mean score of 2.84 for PGY 1–2 (t ⫽ 2.38, df ⫽ 57, P ⬍ 0.05), while PGY3-5 had significantly lower mean laissez-faire score of 0.88 compared with the mean score 0.55 for PGY1-2 (t ⫽ 2.20, df ⫽ 59, P ⬍ 0.05). A profile of the average scores comparing resident gender and PGY profiles are illustrated in Fig. 1. All three leadership outcomes were found to be significantly and positively related with the leadership scales except management by exception-passive and laissez-faire. The five transformational leadership subscales and contingent reward were highly correlated with the three leadership outcome variables (all r ⬎ 0.50). Management by exception-active, the remaining subscale of the transactional leadership dimension, also demonstrated significant and positive correlations with the criterion variables, although the magnitude of the correlations was much weaker that that of contingent reward. Because the two subscales of management by exception-passive and laissez-faire were negatively correlated with the three outcome variables, this supported the existence of a separate passiveavoidant leadership construct. Table 2 provides a full breakdown of the means, standard deviations, reliabilities, and Pearson product-moment correlations between MLQ leadership factors, and outcome variables. Item Factor Structure

A principal component analysis was performed on the item responses to identify if the underlying structures were consistent with the theoretical constructs of transformational theory in the literature. A two factor model emerged, in which the two factors with eigenvalues over 1.0 were 6.47 and 1.43, and 65.86% of the variance was captured with 52.32% in the first factor and 13.55% in the second. Of the 12 components, 10 loaded on a factor comprised of active/effective leadership behaviors, while the other two, which represented passive-avoidant leadership components, management

by exception-passive, and laissez-faire, loaded together as the second factor. This result led to further testing of the leadership models using these two subscales as their own global leadership dimension in the multivariate and hierarchical analyses. Table 3 provides the rotated component matrix showing the loading of items into the two separate factors. Comparison to National Average

Univariate analyses were conducted to determine whether differences existed between the surgical resident sample and an existing normative United States sample (n ⫽ 3375) for all MLQ leadership dimensions [55]. These comparisons were accomplished using onesample t-tests in which the population, or test standards, were the means of the United States normative sample. Three subscales were both found to be significantly different from the United States normative sample beyond the 0.05 level. The management by exception-active was significantly higher than those of the normative sample (t ⫽ 6.6, df ⫽ 58, P ⬍ 0.01). Similarly, the mean of management by exceptionpassive among the resident sample was significantly higher than that of the normative sample (t ⫽ 2.8, df ⫽ 59, P ⬍ 0.01). On the other hand, individualized consideration was significantly lower than the national sample (t ⫽ 2.7, df ⫽ 61, P ⬍ 0.01). Table 4 compares the results between the resident and normative samples. Multivariate and Hierarchical Analysis

Multivariate analyses were subsequently performed to test for potential relationships between the MLQ leadership dimensions and the criterion variables, using the three global leadership dimensions of transformational, transactional, and passive-avoidant leadership, as well as the two demographic variables of gender and PGY. For the outcome variable of extra effort, the model was significant with 64% of the variance in extra effort explained by the three leadership dimensions and the two demographic variables (F ⫽ 17.41, df ⫽ 5, 50, P ⬍ 0.001). Similarly, the models for effectiveness and satisfaction were also significant with a significant portion of the variance accounted for by the specified leadership and demographic variables (R 2 ⫽ 0.56, F ⫽ 12.52, df ⫽ 5, 50, P ⬍ 0.001, and R 2 ⫽ 0.59, F ⫽ 14.23, df ⫽ 5, 50, P ⬍ 0.001, respectively). However, only the transformational leadership dimension was found to be a significant predictor of all of the three outcome variables. Because the multivariate results found the transformational leadership construct to be highly influential on the outcome variables, a series of hierarchical regression analyses was conducted, in which the three leadership outcome variables were used as the dependent variables. This method was essential to test for

HORWITZ ET AL.: LEADERSHIP CHARACTERISTICS OF A SURGICAL RESIDENT COHORT

FIG. 1.

53

Mean Leadership Scores by Gender and PGY.

what has become deemed as the “augmentation effect” in which transformational leadership works to increase the motivation elicited by transactional leadership, hence producing a synergy among leader effectiveness traits [52]. A contingent of leadership theorists have argued that the selective application of transforma-

tional leadership increases the influence of transactional leadership, resulting in improved performance, while also challenging earlier assumptions that transformational and transactional leadership are mutually exclusive constructs [52, 63, 64]. In the current study, analyses were conducted to see whether the augmen-

54

JOURNAL OF SURGICAL RESEARCH: VOL. 148, NO. 1, JULY 2008

TABLE 2 Means, Standard Deviations (SD), and Correlations Among Surgical Resident MLQ Leadership Dimensions and Outcome Variables a

CR IS MBEP MBEA LF II(B) IM II(A) IC EFF EE SAT

Mean

SD

CR

IS

MBEP

MBEA

LF

II(B)

IM

II(A)

IC

EFF

EE

SAT

2.98 2.92 1.30 2.31 .75 2.84 3.05 2.96 2.97 3.11 2.90 3.22

.55 .56 .67 .85 .59 .66 .66 .53 .57 .57 .70 .57

(.68) .60** ⫺.02 .43** ⫺.26 .65** .65** .67** .58** .61** .49** .59**

(.61) ⫺.09 .35** ⫺.11 .68** .73** .55** .62** .63** .62** .58**

(.71) ⫺.04 .45** ⫺.05 ⫺.15 ⫺.11 ⫺.11 ⫺.21 ⫺.18 ⫺.13

(.78) .07 .47** .38** .47** .29* .33* .27* .30*

(.60) ⫺.06 ⫺.16 ⫺.12 ⫺.12 ⫺.21 ⫺.17 ⫺.11

(.71) .74** .71** .58** .64** .57** .67**

(.80) .67** .60** .69** .63** .63**

(.61) .62** .69** .58** .68**

(.65) .68** .66** .68**

(.75) .73** .74**

(.78) .75**

(.57)

CR ⫽ contingent reward; IS ⫽ intellectual stimulation; MBEP ⫽ management by exception-passive; MBEA ⫽ management by exceptionactive; LF ⫽ laissez-faire; II(B) ⫽ idealized influence-behavior; II(A) ⫽ idealized influence-attributed; IC ⫽ individual consideration; EFF ⫽ effectiveness; EE ⫽ extra effort; SAT ⫽ satisfaction. a n ⫽ 64, numbers in parentheses are reliability scores. * P ⬍ 0.05. ** P ⬍ 0.01.

tation effect could be detected in the cohort of surgical residents. Because gender has previously been found to be a potential confounder, as extant research suggested that female leaders tend to be more transformational than their male counterparts [65– 67], and because PGY could also potentially affect self-assessment of leadership style, these two demographics were controlled for in the analyses. In the first step of the regression, the two demographic variables, gender and PGY, were entered, followed by passive-avoidant leadership. Next, transactional leadership was entered. Finally, transformational leadership was introduced to test for an augmentation effect from transformational leadership as would be determined by the finding of a significant change in the R 2 in the model. The significance of the R 2 change caused by the addition of each factor to the model was thus tested, and then the four steps repeated for the three leadership outcome variables that have been found to indicate an augmentation effect: effectiveness, extra effort, and satisfaction [68, 69]. During the first step, only a 0.03 increase in the R 2 was detected when the gender and PGY variables were included in the model, which was not determined to be statistically significant. In the second step, passiveavoidant leadership was added to the model and found to increase the R 2 by 0.07, which was also not found to be a statistically significant predictor of the three outcome variables. In the third step, the addition of transactional leadership to the model resulted in a 0.23 increase in R 2 for effectiveness, which was significant (⌬F ⫽ 16.92, df ⫽ 1, 48, P ⬍ 0.001). Similarly, there was a significant increase of 0.15 in the R 2 of extra effort and 0.26 in the R 2 for satisfaction (⌬F ⫽ 9.99,

df ⫽ 1, 48, P ⬍ 0.01 and ⌬F ⫽ 19.02, df ⫽ 1, 48, P ⬍ 0.001, respectively). Overall, 32% of the total variance in effectiveness was explained by adding transactional leadership to the model, and 23% of the variance for extra effort and 29% for satisfaction were explained by adding transactional leadership to the model. In the final step, the augmentation effect of transformational leadership was further confirmed as there was additional significance in the R 2 change for all three outcome variables as the addition of transformational leadership increased the R 2 for effectiveness by 0.31, by 0.33 for extra effort, and by 0.29 for satisfaction (⌬F ⫽ 43.00, df ⫽ 1, 47, P ⬍ 0.001, .⌬F ⫽ 36.74, df ⫽ 1, 47, P ⬍ 0.001, and .⌬F ⫽ 35.60, df ⫽ 1, 47, P ⬍ 0.001, TABLE 3 Factor Analysis Results for the MLQ Dimensions Component Variable II(B) IM SAT EFF II(A) IS CR IC EE MBEA LF MBEP

1

2

.866 .836 .835 .828 .823 .799 .796 .786 .776 .512 ⫺.094 ⫺.081

.045 ⫺.156 ⫺.124 ⫺.257 ⫺.069 ⫺.060 ⫺.109 ⫺.142 ⫺.279 .189 .846 .805

Note. Extraction method ⫽ Principal component analysis; Rotation method ⫽ Varimax with Kaiser normalization.

55

HORWITZ ET AL.: LEADERSHIP CHARACTERISTICS OF A SURGICAL RESIDENT COHORT

TABLE 4 Comparisons Between the Surgical Resident Averages and the National Normative Sample on the MLQ Dimensions Item CR Resident sample U.S. normative sample IS Resident sample U.S. normative sample MBEP Resident sample U.S. normative sample MBEA Resident sample U.S. normative sample LF Resident sample U.S. normative sample II(B) Resident sample U.S. normative sample II(A) Resident sample U.S. normative sample IM Resident sample U.S. normative sample IC Resident sample U.S.Normative sample EFF Resident sample U.S.normative sample EE Resident sample U.S. normative sample SAT Resident sample U.S. normative sample

n

Mean

SD

t

df

Sig (two-tailed)

59 3375

2.98 2.99

.55

⫺.65

58

⫺.14

59 3375

2.92 2.96

.55

⫺.49

58

.62

60 3375

1.31 1.07

.67

2.75

59

.008*

59 3375

2.31 1.58

.685

6.60

58

.001*

61 3375

.75 .61

.59

1.87

60

.07

62 3375

2.84 2.99

.66

⫺1.77

61

.08

61 3,375

2.96 2.95

.53

⫺.13

60

.89

60 3375

3.05 3.04

.66

⫺.72

59

.87

62 3375

2.97 3.16

.57

⫺2.66

61

.01*

59 3375

3.11 3.14

.57

⫺.40

58

.69

60 3375

2.90 2.79

.70

1.18

59

.24

60 3375

3.22 3.09

.55

1.72

59

.09

* Statistically significant at P ⬍ 0.01.

respectively). The complete results from these analyses are provided in detail in Table 5. DISCUSSION Gender and PGY Differences and Leadership Implications

The leadership item scores between surgical residents by gender, with the exception of management by exception-active, were statistically equivalent. This is true despite past research that has demonstrated different leadership profiles between genders, with greater transformational attributes exhibited by women [65– 67]. The high degree of consistency across the resident leadership profiles may likely be explained by the attraction-similarity-attrition (ASA) hypothesis that individuals with similar personal characteristics are attracted to, and accepted by, similar organizations, with turnover occurring among those that do not

fit well within the group [70, 71]. Tests of the ASA hypothesis have found that one of the primary manifestations of this effect is strong homogeneity among members within the group, and often very congruent values and alignment of goals [72, 73]. Given the long educational process and extensive medical focus preceding the choice to become surgical residents, it is not surprising in light of ASA theory that the leadership styles between residents of different genders would be more similar than divergent. Moreover, the similarity within this sample of residents may have been additionally affected by the unique selection criteria of the surgical residency program which, according to personorganization fit theory [74, 75], would predict that individuals with comparable personality characteristics would be chosen. Very few statistically significant differences were also detected between PGY1-2 and PGY3-5, and for the

56

JOURNAL OF SURGICAL RESEARCH: VOL. 148, NO. 1, JULY 2008

TABLE 5 Results of Hierarchical Regression Analyses for the Augmentation Effects of Transformational Leadership Step

Variable

R 2(adjusted R 2)

1 2 3 4

Gender and PGY PA leadership TC leadership TF leadership

.029 (⫺.008) .096 (.044) .321 (.268) .635 (.599)

1 2 3 4

Gender and PGY PA leadership TC leadership TF leadership

.016 (⫺.021) .079 (.026) .230 (.169) .556 (.512)

1 2 3 4

Gender and PGY PA leadership TC leadership TF leadership

.003 (⫺.034) .030 (⫺.026) .294 (.238) .587 (.546)

⌬R 2

⌬F

F ratio

P

.029 .067 .225 .314

.786 3.875 16.916 43.002

.786 1.844 6.036 17.406

.461 .151 .001 .001

.016 .062 .151 .326

.441 3.518 9.991 36.740

.441 1.481 3.800 12.519

.645 .231 .009 .001

.003 .027 .263 .294

.087 1.444 19.020 35.590

.087 .540 5.300 14.234

.917 .657 .001 .001

Effectiveness

Extra Effort

Satisfaction

aforementioned reasons, this high degree of similarity may be explained. However, one particularly interesting finding was that the PGY3-5 group was significantly lower in laissez-faire leadership and, thus, may adopt more assertive and transformational leadership styles as they gain competence and clinical experience, as leadership effectiveness has been associated with experiential growth and increasing expertise [48]. Additionally, residents may opt to employ more active leadership techniques as they gain rank within the PGY hierarchy due to feeling the need to live up to the role expectations of both the lower PGYs as well as by their attending mentors [76, 77]. The results from comparing the average resident scores to the average scores of the national sample were also quite interesting, as both management by exception-active and management by exceptionpassive were found to be higher than the norm. A possible explanation for the significantly higher divergence on both scores may be that the potential outcomes affecting patient well-being is of such importance that the expectations placed on the residents during their training are extremely high. Management by exception-active techniques may be used when the consequences of a course of action are high and outcomes become foreseeable; management by exceptionpassive may occur when residents either lack the technical expertise to have foresight into the results of an action, or be hesitant to actively manage due to perceived or real doubts about their ability to deal with given situations. Because of the rigid hierarchy of authority in surgical training programs, passive management may actually be expected under well-specified conditions, and thus this style of management may often be appropriate and not indicative of deficient leadership skills. Notably, while the difference between PGY scores was not a predictor of management by exception-passive, but was significant for laissez-

faire leadership, experience and role expectations should not be ruled out as potentially mediating factors. The third finding, that individualized consideration was significantly below the national average, may indicate that the residents are less supportive of each other, and could potentially flow from feelings of high independence coupled with frequent rotation between different surgical services that act to impede the development of close interpersonal relations. Based on such research, it may be shown that intervention to improve active leadership and supportive styles may be desirable, although also possible that the natural progression of experience would allow for a more appropriate transition from passive to active management styles in the case of management by exception leadership. Scale Characteristics and the Augmentation Effect

Consistent with existing research, this study found that transformational leadership scales and contingent reward were highly intercorrelated, suggesting that although discriminant in nature, these subscales converge on the same broad conceptual domain of leadership [78, 79]. This was also supported by the results of the principal components analysis. Although the MLQ is generally held to factor transformational and transactional leadership separately, other studies using this instrument [52, 78, 80] also reported a dichotomous two factor solution with one factor representing an underlying single dimension for transformational and transactional leadership and one factor representing a passive-avoidant leadership dimension. Analysis from the hierarchical regression revealed a strong and significant association between transformational leadership and the three outcome variables of effectiveness, satisfaction, and extra effort. The portion of variance explained by transformational leadership

HORWITZ ET AL.: LEADERSHIP CHARACTERISTICS OF A SURGICAL RESIDENT COHORT

remained significant even when controlling for the two other leadership variables entered on steps two and three. Relative to the other outcome variables, transformational leadership was most strongly associated with high work motivation and satisfaction for the surgeon respondents. Alike past research [81], and on the basis of these results, transformational leadership behaviors appeared to be associated with higher levels of performance and satisfaction, hence validating the augmentation effect in the medical resident sample. These results thus suggest that including transformational style leadership training in surgical educational curriculums would be both worthwhile in itself and compliment the effects of other leadership techniques deemed important by academic medical departments.

57

ever, because the MLQ was able to discern important areas where leadership training was most needed, it strongly suggests that significant benefits for assessing such needs in surgical resident curriculums could be realized through the administration of the instrument as part of designing and improving individual programs. Future research using the MLQ to contrast differences found between different resident specialties, thus identifying both similar and unique results, could also offer important improvements in the design of surgical educational curriculums. REFERENCES 1.

Kamal MF, Itani MD, Liscum K, et al. Physician leadership is a new mandate in surgical training. Am J Surg 2004;187:328.

Study Limitations

2.

Souba WW. The Achilles heels of an academic surgeon: A leadership tale. J Surg Res 2004;123:320.

There are several limitations to this study. First, because the resident identities were blinded, it was not possible to correlate leadership scores to actual performance measures; in this case the performance variables were self-reported measures, and hence potentially subject to self-report bias. Future investigations that would link MLQ results with actual performance measures would be an important extension of this work. Second, because of the unequal distribution of residents by PGY, potentially important growth effects may have been missed due to the lower samples at the higher levels, and necessary dichotomization of these groups. A more robust sample would be desirable for parsing out such potential effects. Third, some important outcomes that would be important to resident performance that have been linked to transformational leadership style, including organizational safety climate [82, 83], creativity processes [84], and optimism [85], were not measured in this study. Future investigations, using the MLQ in conjunction with these potential outcomes, would represent valuable contributions to the surgical leadership literature.

3.

Guthrie MB. Challenges in developing physician leadership and management. Frontiers Health Service Management 1999; 153:45.

4.

McAlearney AS, Fisher D, Heiser K, et al. Developing effective physician leaders: Changing cultures and transforming organizations. Hosp Topics 2005;83:11.

5.

Merry M. Physician leadership. The time is now. Physician Executive 1996;22:4.

6.

Lane DS, Ross V. Defining competencies and performance indicators for physicians in medical management. Am J Prevent Med 1998;14:229.

7.

Yule S, Flin R, Patterson-Brown S, et al. Nontechnical skills for surgeons in the operating room: A review of the literature. Surgery 2006;139:140.

8.

Carthey J, deLeval MR, Wright DJ, et al. Understanding excellence in complex, dynamic medical domains. In: Proceedings of the International Ergonomics Association and Human Factors Society triennial conference. Santa Monica, CA: The Human Factors and Ergonomic Society Press. 2000:136 –139.

9.

Brock M. Leadership qualities in prominent neurosurgeons. Act Neurochir 1977;69:8.

10.

DeJong SA. Redefining success as significance through leadership in the new economy. Am J Surg 2006;191:291.

11.

Schwartz RW. Physician leadership: A new imperative for surgical educators. Am J Surg 1998;176:38.

12.

Awad SS, Hayley B, Fagan SP, et al. The impact of a novel resident training curriculum. Am J Surg 2004;188:481.

13.

McKnight L, Stetson PD, Bakken S, et al. Perceived information needs and communication difficulties of inpatient physicians and nurses. Washington, D.C.: Proceedings of the AMIA Symposium, November 2001:453– 457.

14.

Yun S, Faraj S, Simms HP Jr. Contingent leadership and effectiveness of trauma resuscitation teams. J Applied Psychol 2005;90:1288.

15.

Friedman DM, Berger DL. Improving team structure and communication; a key to hospital efficiency. Arch Surg 2004;139: 1194.

16.

Awad SS, Fagan SP, Bellows C, et al. Bridging the communication gap in the operating room with medical team training. Am J Surg 2005;190:770.

17.

Yun S, Faraj S, Xiao Y, et al. Team leadership and coordination in trauma resuscitation. Advances in Interdisciplinary Study of Work Teams 2003;9:189.

CONCLUSIONS AND DIRECTIONS FOR FUTURE RESEARCH

This study was the first to use a validated psychometric instrument to assess the leadership skills of surgical residents during their training process. In doing so, both positive and negative leadership styles were identified. The use of the MLQ thus provided valuable information for identifying areas for future surgical training. It is important to note that because the subjects examined in this investigation were surgical residents from only a single cohort, they may represent a sample that shares characteristics that are uniquely similar to this resident group, and hence limit the degree in which the results can be generalized to residents in other surgical education programs. How-

58 18.

19.

20. 21.

22.

23.

24. 25.

26.

27.

28. 29.

30. 31. 32. 33. 34.

35. 36. 37. 38. 39. 40. 41. 42.

JOURNAL OF SURGICAL RESEARCH: VOL. 148, NO. 1, JULY 2008 Manias E, Street A. The interplay of knowledge and decision making between nurses and doctors in critical care. Int J Nurs Stud 2001;38:129. Sutcliffe KM, Lewton E, Rosenthal MM. Communication failures: An insidious contributor to medical mishaps. Acad Med 2004;79:186. Dwyer K. Communication breakdowns combine to cause system failures. Int J Qual Health Care 2001;13:265. Lingard L, Espin S, Whyte S, et al. Communication failures in the operating room: An observational classification of recurrent types and effects. Qual Saf Health Care 2004;13:330. Risser DT, Rice MM, Salisbury ML, et al. The potential for improved teamwork to reduce medical errors in the emergency department. Ann Emerg Med 1999;34:373. Taylor-Adams S, Vincent C. Clinical accident analysis: Understanding the interactions between the task, individual, team, and organization. In: Vincent C, de Mol B, Eds. Safety in Medicine. Amsterdam: Pergamon Press, 2000:101–116. Ong LM, de Haes JC, Hoos AM, et al. Doctor-patient communication: A review of the literature. Soc Sci Med 1995;40:903. Smith VA, DeVellis BM, Kalet A, et al. Encouraging patient adherence: Primary care physicians’ use of verbal compliancegaining strategies in medical interventions. Patient Educ Counsel 2005;57:62. Levinson W, Chaumenton N. Communication between surgeons and patients in routine office visits. Surgery 1999;125:2: 127. Schneider J, Kaplan SH, Greenfield S, et al. Better physician– patient relationships are associated with higher reported adherence to antiretroviral therapy in patients with HIV infection. J Gen Int Med 2004;19:1096. Stewart MA. Effective physician–patient communication and health outcomes: A review. Can Med Assn J 1995;152:1423. Thomas EJ, Sexton JB, Helmreich RL. Discrepant attitudes about teamwork among critical care nurses and physicians. Crit Care Med 2003;31:992. Gilinas L, Loh DY-H. The effect of workforce issues on patient safety. Nursing Econ 2004;22:266. Souba WW, Day DV. Leadership values in academic medicine. Acad Med 2006;81:20. Buryak JS. Can physicians lead other physicians into the future? Physician Executive Sept.–Oct. 1998. Stogdill RM. Personal factors associated with leadership: A review of the literature. J Psychol 1948;25:35. French JRP, Raven BH. The basis of social power. In: D Cartwright, A Zandler, Eds. Group Dynamics: Research and Theory, 2nd ed. New York: Harper Row, 1960;607– 623. Tannenbaum R, Schmidt WH. How to choose a leadership pattern. Harvard Business Rev, March/April 1957. Blake RR, Mouton JS. The managerial grid. Houston, TX: Gulf, 1964. Fiedler FE. A theory of leadership effectiveness. New York: McGraw-Hill, 1967. House RJ. A path-goal theory of leader effectiveness. Admin Sci Quarterly 1971;16:321. Hersey P, Blanchard KH. Life-cycle of leadership. Training and development J May, 1969. Hersey P, Blanchard H. Management of organizational behavior, 4th ed. Englewood Cliffs, NJ: Prentice Hall, 1982. Argyris C. Interpersonal competence and organizational effectiveness. Homewood, IL: Irwin Dorsey Press, 1976. Argyris C. Personality and organizations. New York: Harper and Row, 1957.

43. 44. 45.

46.

47. 48. 49.

50.

51. 52. 53. 54. 55. 56. 57.

58.

59.

60.

61.

62. 63.

64.

65.

66.

Kelly RE. In praise of followers. Harvard Business Rev 1988; 55:67. Goleman D. Emotional intelligence. New York: Bantam Books, 1995. Mayer JD, Salovey PR, Caruso DR. Emotional intelligence meets traditional standards for an intelligence. Intelligence 1999;27:267. Stratton TD, Elam CL, Murray-Spenser AE, et al. Emotional intelligence and clinical skills: Preliminary results from a comprehensive clinical performance evaluation. Acad Med 2005;80: 534. A critique of emotional intelligence. Murphy KR, ed. Mahwah, NJ: Lawrence Erlbaum Associates, 2006. Hughes RL, Ginnett RC, Curphy GJ. Leadership: Enhancing the lessons of experience, 2nd ed. Chicago: Irwin, 1996. Yukl G, Van Fleet DD. Theory and research on leadership in organizations. In: Dunnette MD, Hough LM, Eds. Handbook of Industrial and Organizational Psychology, Vol III. Palo Alto, CA: Consulting Psychological Press, 1992:147–198. Roach CF, Behling O. Functionalism: Basis for an alternative approach to the study of leadership. In: Hunt JG, Hosking DM, Shreisheim CA, Steward R, Eds. Leaders and Managers: International perspectives on managerial behavior and leadership. Elmsford, NY: Pergamon Press, 1984:51– 61. Burns JM. Leadership. New York: Harper and Row, 1978. Bass BM. Leadership and Performance Beyond Expectations. New York: Free Press, 1985. Bass BM. From transactional to transformational leadership: Learning to share the vision. Org Dynamics 1990;18:19. Bass BM, Riggio RE. Transformational leadership, 2nd ed. Mahwah, NJ: Lawrence Erlbaum Associates, 2006. Avilio BJ, Bass BM. Multifactor Leadership Questionnaire, 3rd ed. Menlo Park, CA: Mind Garden Inc., 2004. Hunt JG. Leadership: A new synthesis. Newbury Park, CA: Sage, 1991. Kanste O, Miettunen J, Kyngas H. Psychometric properties of the Multifactor Leadership Questionnaire among nurses. J Adv Nurs 2007;57:201. Eid J, Johnson BH, Brun W, et al. Situation awareness and transformational leadership in senior military leaders: An exploratory study. Mil Psychol 2004;16:203. Dvir T, Eden D, Avolio BJ, et al. Impact of transformational leadership on follower development and performance: A field experiment. Acad Management J 2002;45:735. Molero R, Morales JF. A study in leadership in health care settings using Bass’ Multifactor Leadership Questionnaire (MLQ). Paper presented at the International Congress of Applied Psychology. Madrid, Spain, July 1994. Xirasagar S, Samuels ME, Stoskopf CH. Physician leadership styles and effectiveness: An empirical study. Med Care Res Rev 2005;62:720. Organ DW. Organizational citizenship behavior: The good soldier syndrome. Lexington, MA: Lexington Books, 1988. Hater JJ, Bass BM. Superiors’ evaluations and subordinates perceptions of transformational and transactional leadership. J Applied Psychol 1988;73:695. Yammarino FJ, Spangler WD, Bass BM. Transformational leadership and performance: A longitudinal investigation. Leadership Quarterly 1993;4:81. Bass BM, Avilio BJ, Atwater L. The transformational and transactional leadership of men and women. Applied Psychol. [An International Review] 1996;45:5. Eagly AH, Johannesen-Schmidtt MC, van Engen ML. Trans-

HORWITZ ET AL.: LEADERSHIP CHARACTERISTICS OF A SURGICAL RESIDENT COHORT formational, transactional, and laissez-faire leadership styles: A meta-analysis comparing men and women. Psychol Bulletin 2003;129:569. 67.

Walumbwa FO, Wu C, Ojode L. Gender and instructional outcomes: The mediating effects of leadership styles. J Management Dev 2004;23:2.

68.

Seltzer J, Bass BM. Transformational leadership: Beyond initiation and consideration. J Management 1990;16:693.

69.

Howell JM, Avolio BJ. Transformational leadership, transactional leadership, locus of control, and support for innovation: Key predictors of consolidated business-unit performance. J Applied Psychol 1993;78:891.

70.

Schneider B. The people make the place. Personnel Psychol 1987;40:437.

71.

Jackson SE, Brett JF, Sessa VI, et al. Some differences make a difference: Individual dissimilarity and group heterogeneity as correlates of recruitment, promotion, and turnover. J Applied Psychol 1991;76:675.

72.

Schneider B, Goldstein HW, Smith DB. The A-S-A framework; an update. Personnel Psychol 1995;48:747.

73.

Schneider B, Smith DB, Taylor S, et al. Personality and organizations: A test of the homogeneity of personality hypothesis. J Applied Psychol 1998;83:462.

74.

French JRP. Person-role fit. Occup Mental Health 1973;3:15.

75.

Rynes SL. Recruitment, job choice, and post-hire consequences: A call for new research directions. In: Dunette MD, Hough LM, Eds. Handbook of Industrial and Organizational Psychology, Vol II. Palo-Alto, CA: Consulting Psychology Press, Inc., 1991: 399 – 444.

76. 77.

78.

79.

80.

81.

82.

83.

84.

85.

59

Hogg MA. A social identity theory of leadership. Personality Social Psychol Rev 2001;5:184. Messick DM. On the psychological exchange between leaders and followers. In: DM Messick, RM Kramer, Eds. The Psychology of Leadership. Mahwah, NJ: Lawrence Erlbaum Associates, 2005:81–96. Bycio P, Hackett RD, Allen JS. Further assessments of Bass’ conceptualization of transactional and transformational leadership. J Applied Psychol 1995;80:468. Lowe KB, Kroeck G, Sivasubramaniam N. Effectiveness correlates of transformational and transactional leadership: A metaanalytic review. Leadership Quarterly 1996;1:385. Waldman DA, Bass BM, Einstein WO. Leadership and outcomes of performance appraisal processes. J Occup Psychol 1987;60:177. Waldman DA, Bass BM, Yammarino FJ. Adding to contingentreward behavior: The augmenting effect of charismatic leadership. Group Org Studies 1990;15:381. Barling J, Loughlin C, Kelloway EK. Development and test of a model linking safety-specific transformational leadership and occupational safety. J Applied Psychol 2002;87:488. Zohar D. The effects of leadership dimensions, safety climate, and assigned priorities on minor work injuries in work groups. J Org Behavior 2002;23:75–92. Kahai SS, Sosik J, Avolio BJ. Effects of leadership style and rewards on creativity-relevant processes and outcomes in an electronic meeting system context. Leadership Quarterly 2003; 14:499. McColl-Kennedy JR, Anderson RD. Impact of leadership style and emotions on subordinate performance. Leadership Quarterly 2002;13:545.