Home Business & Finance Distance Leadership in International Corporations: Why Organizations Struggle when Distances Grow
Direct Effects of Leadership Behavior on Follower Self-Leadership and Performance
The first sequence of hypotheses is concerned with the effects of facets of Full Range Leadership on work-related outcomes. Work-related outcomes are specified as self-leadership and individual performance. The research framework is displayed in Figure 6.
Prior to computing regression coefficients, the intercorrelation matrix was used to illustrate potential relations between variables. Interpretation of intercorrelatedness and magnitude of coefficients have long been subject to discussion. A widely used framework is provided by Cohen (1988, pp. 79-80) who suggests that coefficients
© Springer Fachmedien Wiesbaden 2017
N. Poser, Distance Leadership in International Corporations,
Advances in Information Systems and Business Engineering,
at .10 are small, whereas .30 are medium, and .50 and above are large. According to a study by Hemphill (2003), the magnitude of correlation coefficients in psychology research must be differentiated as the assumption of what is perceived as high is estimated upon different sets of empirical guidelines. The framework is based upon a review of 380 meta-analytic studies, disclosing that one third of the studies show correlation magnitudes below .20, one third between .20 and .30, and one third above .30 (Hemphill, 2003, p. 78).
Figure 6. Influences of Leadership Behavior on Work-Related Outcomes
Looking at the intercorrelation matrix in Table 17, relations between perceptions of Full Range Leadership, self-leadership, and performance reveal high interdimension correlations. Transactional leadership demonstrates a strong and significant correlation (r = .71, p < .001) with transformational leadership. Detailed reviews produce strong positive correlations with contingent reward (r = .87, p < .001) and moderate correlations with MBEa (r = .30, p < .001). As expected, passive leadership (r = -.64, p < .001) together with MBEp (r = -.25, p < .001), and laissez-faire (r = -.73, p < .001) show strong significant negative correlations with transformational leadership.
Similarly, transactional leadership reports significant negative correlation with the higher-order factor of passive leadership (r = -.34, p < .001). The decision to assume passive MBE to be serving as subfacet of passive leadership rather than transactional leadership is reassured, as passive MBE does not show intercorrelations with transactional leadership (r = .04, n.s.) but shows a strong interrelation to the higher-order factor of passive leadership (r = .76, p < .001). Laissez-faire leadership confirms assumptions to result in a statistically significant negative association when related to transactional behavior (r = -.56, p < .001).
Observing the relations between self-leadership and dimensions of the FRL, all facets but for one display statistically significant results. The correlation between transformational leadership and self-leadership indicates significant positive direction at a high significance level (r = .24, p < .001). All subfacets of FRL - attributed idealized influence (r = .18, p < .001), idealized influence (behavior) (r = .29, p < .001), inspirational motivation (r = .25, p < .001), intellectual stimulation (r = .21, p < .001), and individualized consideration (r = .15, p < .01) - show positive correlations with self-leadership of low to moderate strength. For transactional leadership, a similar picture is illustrated. The higher-order factor of transactional leadership expresses positive relationship with self-leadership (r = .29, p < .001). Examining the subfacets more closely, contingent reward (r = .20, p < .001) and active management-by-exception (r = .27, p < .001) both correlate significantly with selfleadership. Passive leadership is found to be the only higher-order factor that does not directly relate to self-leadership (r = -.06, n.s.). However, the subfacet passive MBE (r = .17, p < .001) reveals positive correlation, whereas laissez-faire leadership (r = -.26, p < .001) results in negative association with self-leadership.
For correlations of leadership behavior with individual follower performance, the higher-order factor of transformational leadership shows weak positive relation (r = .16, p < .01). Idealized influence (attributed; behavior) (r = .20, p < .001; r = .12, p < .05), inspirational motivation (r = .19, p < .001), and individualized consideration (r = .11, p < .05) show weak to medium statistically significances in relation to follower performance. For facets of transactional leadership, only contingent reward leadership (r = .12, p < .01) shows significant positive correlation with follower performance. Passive leadership as higher-order factor correlates negatively with follower performance (r = -.17, p < .001), with a marginally greater influence indicated by the subfacet of laissez-faire (r = -.14, p < .01) than passive MBE (r = - .12, p < .05). Assessing the correlation between self-leadership and individual follower performance, the matrix points out that there might be indications for direct positive relation of medium strength (r = .24, p ^ .001).
Note, n = 372. * p < .05. ** p < .01, *** p < .001, TF = transformational leadership, Ha = idealized influence (attributed), lib = idealized influence (behavior), IM = inspirational motivation, IS = intellectual stimulation, IC = individualized consideration, TK = transactional leadership, CR = contingent reward, MBEa = active management-by-exception, PL = passive leadership, MBEp = passive management-by-exception, LF = laissez-faire, SL = global self-leadership, Perf = performance
Table 18 outlines correlations between perceptions of FRL higher-order factors, self-leadership, performance, and control variables. Leader gender, age, and educational background do not display any significant correlations. Leadership experience and age (r = .67, p ^ .001) show high interrelatedness as older organizational members tend to have gained more leadership experience. Transformational leadership is related to leaders’ level of hierarchy (r = .22, p < .05). Higher levels of hierarchy accompany an increase in transformational leadership behavior. Also, hierarchy level and leadership experience are positively correlated (r = .25, p < .01). Functional area displays positive correlations with transformational (r = .39, p ^ .001) and transactional leadership (r = .25, p < .01), self-leadership (r = .31, p ^ .001), and negative correlation with passive leadership (r = -.27, p < .05). The positive linkages explain that, in indirect/administrative areas, transformational leadership, transactional leadership, and self-leadership are more distinctive. Passive leadership is yet lower in these departments. Furthermore, male followers rated their leaders’ behavior as more transactional (r = .12, p < .05).
Tenure with leader shows weak positive correlation with transformational leadership (r = .11, p < .05) and transactional leadership (r = .15, p < .01). The table also reveals that tenure with leader is more likely to be higher in indirect/administrative areas (r = .19, p < .05), and older employees tend to stay longer with one leader (r = .24, p ^ .001).
Table 18. Intercorrelations of FRL, Self-Leadership, Performance, and Control Variables
Note, n = 372. * p < .05. ** p < .01, *** p < .001, TF = transformational leadership, TK = transactional leadership, PL = passive leadership, SL = global self-leadership, Perf = performance, L Gen = leader gender, L Age = leader age, L Exp = leadership experience, L Edu = leader education, L Hier = leader hierarchy level, Area = functional area, F Gen = follower gender, F Age = follower age, F Ten = follower tenure.
Transformational leadership and transactional leadership behavior both predict positive follower self-leadership, whereas passive leadership behavior predicts negative follower self-leadership.
Multiple linear regressions are computed to examine effects of each individual dimension on follower self-leadership. The calculated model reveals R2 = .092. In other words, 9.2% of follower self-leadership is explained by perceived FRL behaviors. Buhner and Ziegler (2009, p. 663) consider these to be low to medium effects. Standardized beta values of regressions are reported in Figure 7.
Figure 7. Influences of Leadership Behavior on Self-Leadership
Note. L = leader, F = follower
Analyzing the impact of transformational, transactional, and passive leadership on follower self-leadership, regressions reveal positive effects only for transactional leadership (P = .21, p < .01). Higher-order factors of transformational and passive leadership do not provide a significant indication of direct relation to selfleadership. Looking at the nine subfacets of FRL, the results show a more differentiated picture. Idealized influence (behavior) (P = .25, p < .01), inspirational motivation (P = .21, p < .05), MBEa (P = .13, p < .05), and MBEp (P = .17, p < .001) show significant positive effects, whereas laissez-faire leadership (P = -.17, p < .05) demonstrates negative influence on follower self-leadership. Hypothesis 1.1 should thus be partially rejected as transformational leadership was projected to influence self-leadership positively and passive leadership to influence self-leadership negatively. Standardized beta weights of FRL subfacets and their effects are displayed in Table 19.
Table 19. Predicting Self-Leadership by FRL Subfacets
Note. n = 372. * p < .05, ** p < .01, *** p < .001; L = leader, F = follower
Effects of control variables
Significances could be detected for none of the regressed control variables, except for functional area ф = .29, p < .05). This result raises the assumption that whether one works in supply chain or in indirect/administrative areas influences the effect of perceived leadership behavior on follower self-leadership. The next step thus includes a t-test to determine whether the differences between groups are significant. The t-test compares mean scores of followers working in manufactur- ing/logistics/supply chain (M = 3.13, SD = 0.66) with those working in indirect ar- eas/administration (M = 3.52, SD = 0.54). The outcome of the t-test suggests that followers working in indirect/administration areas ascribe themselves a considerably higher level of self-leadership compared to those working in manufactur?ing/logistics or supply chain (t = -3.54, df = 120, p < .001). Influence of the dichotomous variable is further tested being split in (1) all cases indicating to be working in supply chain, and (2) all cases working in indirect/administrative areas. Computations reveal that transactional leadership does not predict self-leadership for individuals working in indirect/administrative areas (P = .16, t = 1.17, n.s.). For subordinates working in supply chain, transactional leadership predicts follower selfleadership (P = .52, t = 2.06, p < .05).
Transformational leadership and transactional leadership behavior both predict positive follower performance, whereas passive leadership behavior predicts negative follower performance.
The second hypothesis concerns the effects of leadership behavior on individual follower performance. According to the minimal values outlined in the correlation matrix, direct effects of perceived leadership behavior on follower performance are expected to be low. Multiple linear regression modeling reveals that neither transformational, nor transactional, nor passive leadership show any significant direct effects on individual performance. Nor does any control variable significantly influence this relationship. Hypothesis 1.2 is hence rejected. Standardized beta weights of regressions are displayed in Figure 8.
Figure 8. Influences of Leadership Behavior on Performance
Note. L = leader, F = follower
Similar to in the prior procedure, subfacets of FRL are regressed to see whether any of the nine behaviors predicts individual performance in particular. Creating the model with subfacets, the value of R squared (R2 = .065) indicates that only 6.5% of individual performance is explained directly by perceptions of leadership behavior. Computing standardized beta coefficients, only one subfacet (attributed idealized influence) returns significant, albeit limited, results. Table 20 releases standardized beta weights and p-values of the regressions.
Table 20. Predicting Performance by FRL Subfacets
Note. n = 372. * p < .05, ** p <. 01, *** p < .001; L = leader, F = follower
The present study examined additional variance that is explained by transformational leadership over the effects of transactional leadership on (1) performance and (2) self-leadership. The first problem occurred when transactional leadership was not found to be significantly directly related to followers’ individual performance (R2 = .006, n.s.). However, when transformational leadership was added to the re?gression, a direct positive effect was reported and the explanation of variance became significant (R2 = .026, F = 4.97, p < .01). This finding could be confirmed by numerous previous studies (Elenkov, 2002; Howell & Avolio, 1993; Waldman, Bass & Yammarino, 1990). Although transactional leadership explained a significant variance of the contribution to followers’ self-leadership, the increase of variance that is added by transformational leadership is marginal (A R2 = .002).
Follower self-leadership strategies have a direct positive effect on follower performance.
To retain insights on which strategies in particular influence follower performance, higher-order factors of self-leadership are analyzed. Standardized beta weights and p-values are provided in Table 21. The examination of self-leadership higher-order factors reveals that cognition-based strategies do not contribute to predict individuals’ performance (P = .00, n.s.). Yet natural reward strategies (P = .15, p < .05) as well as social self-leadership strategies (P = .14, p < .05) indicate weak statistically significant effects on performance.
As influences of leadership behavior on performance show weak direct effects only for attributed idealized influence, yet direct effects of self-leadership on performance indicate significant increases of natural-reward strategies and social selfleadership strategies, self-leadership is controlled for as moderator in the relation between leadership behavior and follower performance.
Three different models are calculated, one for each higher-order leadership behavior. For testing moderation effects, predictor variables are centralized. Predictor and potential moderator are subsequently expressed as product term and regressed, together with the original predictor on the outcome variable. If the product of predictor and potential moderator becomes statistically significant, a moderation effect occurs (Buhner & Ziegler, 2009, pp. 725ff).
For the first model, assessing the potential of self-leadership to moderate the relation between transformational leadership and follower performance (R2 = .068), the result of the regression (P = .01, n.s.) does not support the existence of moderation effects. The second model aims at testing moderating influences of selfleadership on the relation between transactional leadership and performance. Predictors are transformed and consequently multiplied. Linear regression computa?tions of the second model (R2 = .06) do not reveal any significant moderation effects (P = .00, n.s.). Testing potential moderation effects of self-leadership on the relationship between passive leadership and performance (R2 = .082), no statistically significant beta-coefficients are retrieved (P = .01, n.s.). To summarize the findings outlined in this paragraph: self-leadership does not act as moderator in the relationship between perceptions of FRL and follower performance.
Table 21. Predicting Performance by Higher-Order Factors of Self-Leadership
Note. n = 372. * p < .05, ** p < .01, *** p < .001
To test the direct linear relation between follower self-leadership and performance standardized beta coefficients are calculated. The model fit index indicates that 5.8% of variance is explained by self-leadership (R2 = .058). Results of the regression function reveal that self-leadership strategies do have a direct statistically significant effect on performance (P = .24, p < .001). Hypothesis 1.3 is thus accepted.
|< Prev||CONTENTS||Next >|