KM practices as mediating variables – hypothesis 2

To test, if hypothesis two: IC assets have a positive impact on organizational performance, when being mediated by KM practices, was lended support, organizational performance was regressed on the variables. Nine sub-hypotheses were tested.

Hypothesis 2a) Human capital -> KM practices -> Organizational performance tested

Human capital had a strong, direct impact (β = 0.40) on organizational performance (Table 6.). The impact of human capital on the mediators is shown in Appendix 5.

When KM practices were regressed on human capital, they got values of: leadership (β = 0.67), development (β = 0.63), and compensation (β = 0.51). The impact of human capital on KM practices mediators was strong and significant (P = .0001).

When these three KM practices mediators were added to the regression analysis one by one, the impact of human capital on value was weaker (Figure 13.) than the direct impact. Instead of showing an effect of β = 0.40, the impact of human capital on value mediated by leadership was β = 0.19, development β = 0.32, and compensation β = 0.35. The impact of human capital on value was less significant when leadership (P = .0035) functioned as the mediator.

Analysis 10 Analysis 11

Analysis 12 Figure 13. Human capital mediated by KM practices

The values of the parameter estimates were compared by calculating: human capital without a mediator (direct impact) – human capital with a mediator = indirect effect, and the results were:

10) Human capital -> leadership -> value: 0.40 – 0.19 = 0.21 11) Human capital -> development -> value: 0.40 – 0.32 = 0.08 12) Human capital -> compensation -> value: 0.40 – 0.35 = 0.05

The impacts of development and compensation on value, were not significant in the model. Therefore the results implicated by these variables need to be critically observed. If they are excluded from the mediation point of view, the only KM practice that mediated human capital was leadership (0.21). The coefficients of determination of the models were good: 10) 27,9 %, 11) 21,4 % and 12) 22,5 % (adjusted), and the t-tests measuring the goodness of fit were also on an good level:

leadership 5,88; development 3,46 and compensation 3,71.

The findings implicated that human capital did have a positive impact on value, whether being mediated by KM practices or not. As in the previous analyses, full mediation was far from being confimed, and partial mediation was found instead when both; the size of the coefficients and the significance of the variables, was analyzed. Sub-hypothesis 10 got supported. The significance of the effect of human capital on value got weaker when leadership functioned as a mediator. This supported partial mediation. What was interesting, was that leadership had a very big contribution (0.21) on the overall impact of human capital on value. The findings suggest, is that 52,5 % of the impact (0.40) of human capital on value, may actually be mediated by leadership. Because the significance of the effect of human capital on value did not weaken, when development and compensation functioned as mediators, mediation was not supported in these other sub-hypotheses.

Hypothesis 2b) Structural capital -> KM practices -> Organizational performance tested

Structural capital had a strong, positive impact (β = 0.25) on organizational performance (Table 6.). The impact of structural capital on the mediators is shown in Appendix 5. When KM practices were regressed on structural capital, they got values of: leadership (β = 0.42), development (β = 0.44), and compensation (β = 0.27). The impact of structural capital on the mediators was significant (P = .0001), except for the impact on compensation that had a bit weaker significance (P = .0033).

When these KM practices mediators were introduced to the regression analysis one by one, the impact of structural capital on value got weaker (Figure 14.). Instead of showing an effect of the exact same size as the direct impact was (β = 0.25), structural capiral got values of β = 0.10 (leadership as a mediator), β = 0.17 (development as a mediator), and β = 0.21 (compensation as a mediator). The impact of structural capital on value was less significant when leadership (P = .0312) or development (P = .0003) functioned as the mediator.

Analysis 13 Analysis 14

Analysis 15 Figure 14. Structural capital mediated by KM practices

When the values of the parameter estimates were compared by calculating:

structural capital without a mediator (direct impact) – structural capital with a mediator = indirect effect, the results were:

13) Structural capital -> leadership -> value: 0.25 – 0.10 = 0.15 14) Structural capital -> development -> value: 0.25 – 0.17 = 0.08 15) Structural capital -> compensation -> value: 0.25 – 0.21 = 0.04

There was positive mediation supported by the analyses. According to the analyses 0.04 to 0.15 of the effect of structural capital on value was explained by a KM practices mediator. Structural capital did not have a significant impact on compensation, which indicated that there might be other factors impacting the relationship. The coefficients of determination of the models were good: 13) 26,2 %, 14) 17,1 % and 15) 18,6 % (adjusted), and the t-tests measuring the goodness of fit were also on an good level: leadership 7,01; development 4,24 and compensation 4,72.

These findings implicated that structural capital did have a positive impact on value, whether being mediated by KM assets or not. Full mediation was far from being

confimed, but partial mediation was found. Sub-hypothesis 13 and 14 got supported.

The significance of the effect of structural capital on organizational value got weaker when either leadership or development functioned as a mediator. This supported partial mediation. What was interesting, was that leadership functioned as such a strong mediator that it accounted for over half of the total impact of structural capital on organizational performance. This finding suggests that leadership may actually mediate 60 % (0.15 of the assumed direct impact 0.25) of the impact of structural capital on value. Compensation did not mediate the impact of structural capital on value, because the significance of the impact did not change when compensation was added to the model.

Hypothesis 2c) Relational capital -> KM practices -> Organizational performance tested

Relational capital had a strong, direct impact (β = 0.37) on organizational performance (Table 6.). The impact of relational capital on the mediators is shown in Appendix 5. When KM practices were regressed on relational capital, they got values of: leadership (β = 0.42), development (β = 0.49), and compensation (β = 0.28). The impact of relational capital on the mediators was significant (P = .0001), except for the impact on compensation that had a bit weaker significance (P = .0107).

When these KM practices mediators were added to the regression analysis one by one, the impact of the relational capital on value got weaker (Figure 15.). The impact of relational capital on value mediated by leadership was β = 0.24, development β

= 0.31, and compensation β = 0.34. The impact of relational capital on organizational performance remained significant (P = .0001).

Analysis 16 Analysis 17

Analysis 18 Figure 15. Relational capital mediated by KM practices

When the values of the parameter estimates were compared by calculating:

relational capital without a mediator (direct impact) – relational capital with a mediator = indirect effect, the results were:

16) Relational capital -> leadership -> value: 0.37 – 0.24 = 0.13 17) Relational capital -> development -> value: 0.37 – 0.31 = 0.06 18) Relational capital -> compensation -> value: 0.37 – 0.34 = 0.03

According to the analyses 0.03 to 0.13 of the effect of relational capital on value was explained by a KM practices mediator. Relational capital did not have a significant impact on compensation, which indicated that there might be other factors impacting this relationship. The impact of development on the value was less significant in the model. If sub-hypotheses addressing compensation and development were excluded from the analysis, because of these findings, leadership would be the only possible mediator. Because there were no changes in the significance of the effect of relational capital on organizational performance when leadership functioned as the mediator, mediation was not supported in these sub-hypotheses. The coefficients of determination of the models were good however: 16) 32,2 %, 17) 24,1

% and 18) 25,4 % (adjusted), and the t-tests measuring the goodness of fit were also on an good level: leadership 6,61; development 3,52 and compensation 4,26.

These findings implicated that relational capital did have a positive impact on value, whether being mediated by KM practices or not. Mediation was not supported, but it looks like there were other factors that impacted the relationship between relational capital and organizational performance. The significance of impact of relational capital on value did not change after introducing KM practices as mediators.

To summarize hypothesis two

IC assets have a positive impact on organizational performance, when being mediated by KM practices, got supported (Appendix 7). None of the models were able to show full mediation, but sub-hypothesis 10, 13 and 14 showed partial mediation. The highest change in the size of the coefficient (0.21) was shown in sub-hypothesis 10. The size of the coefficient in sub-hypothesis 13 changed 0.15 and 0.08 in sub-hypothesis 14. The biggest change in the significance (.0312) of the predictor variable was found in sub-hypothesis 13. The significance of the predictor in sub-hypothesis ten was .0035 and .0003 sub-hypothesis 14. These findings make mediation practically non-existent, because the significance of the effects did not change substantially. The strongest support for mediation was found in the effect of

structural capital that dropped from 0.0001 to 0.312 when mediated by leadership (sub-hypothesis 13). One thing to point out is that leadership was the strongest mediator of the analyses. Even though structural capital showed the biggest change in the significance of the impact when mediated by leadership, the biggest change in the absolute size of the coefficient was that of human capital when mediated by leadership. When only the coefficients were compared, there were direct and indirect effects found in everyone of the nine sub-hypothesis.