Page - (000235) - in Knowledge and Networks

230 Model 1 in Table 11.2 shows that each of the three control variables (individual’s rank, tenure in the organization, and job performance) significantly predicted the extent to which the individual was perceived as a leader. Recall that the first hypoth- esis suggested that the higher the self-monitoring score, the more the individual would be perceived as a leader. We found support for this prediction in the regres- sion analysis summarized in Model 2, which shows that the addition of self- monitoring significantly improved model fit over the controls-only Model 1 (χ = 7.86, p < .01). Thus, high self-monitors tend to be perceived as leaders. But are they also per- ceived as advice providers as suggested by Hypothesis 2? The answer is yes, as shown by the analyses summarized in Table 11.3. Controlling for the individual’s rank, tenure, and job performance, the addition of self-monitoring in Model 2 sig- nificantly improved model fit over the base-line Model 1 (χ = 4.80, p < .05). More co-workers report that high self-monitors, relative to low self-monitors, provide them with advice. Hypothesis 3 suggested that the self-monitoring scores of those individuals who were trusted by people who distrusted each other would predict emergent leader- ship. We found support for this trust-leveraging hypothesis. As shown in Table 11.2, Model 5, the interaction between self-monitoring and trust brokerage was signifi- cant (p < .01). The inclusion of the interaction term improved model fit relative to the main-effects Model 4 (χ = 5.32, p < .05), suggesting that high self-monitors, rela- tive to low self-monitors, leveraged go-between positions in the trust network to emerge as leaders in the eyes of others. To more closely examine the form of these interactions, we constructed a split plot following the procedures described in Aiken and West (1991). The form of the plot depicted in Fig. 11.1 shows support for Hypothesis 3: although trust brokerage was positively related to the number of leadership nominations received for both high self-monitors and low self-monitors, this relationship was significantly stron- ger for highs than for lows. Table 11.1 Means, standard deviations, and correlations of the variables Variable M SD 1 2 3 4 5 6 1. Rank 0.22 0.42 2. Tenure 53.95 39.25 .19 3. Job performance 10.15 2.75 .25* .28** 4. Self- monitoring 0.07 0.08 .18+ 5. Trust brokerage 0.29** 0.14 .08 .20+ 6. Perceived leadership 0.72*** 0.29 .23* .25** .35*** .23* .36** 7. Advice centrality 15.87 11.11 .60*** .28** .23* .25* .52*** .74*** Note +p < .10; *p < .05; **p < .01; ***p < .001 M. Kilduff et al.