CHAPTER 4 DATA ANALYSIS AND RESULTS
4.4 Regression Analysis …
4.4.3 The Effect of Organizational Culture on Knowledge Creation Modes
There are three regressions. First, IVs are MC, AC, HC, and CC and DV is G-DR.
Second, IVs are MC, AC, HC, and CC and DV is G-FR. Third, IVs are MC, AC, HC, and CC and DV is G-DE. For hypothesis 3, Table 4-23 indicated that adhocracy culture has a positive and significant influence on goal-free creation modes (β = 0.182,
p-value = 0.072). Market culture (β = 0.032, p-value = 0.032), adhocracy culture (β =
0.166, p-value = 0.041), hierarchy culture (β = 0.247, p-value = 0.024), and clan culture (β = 0.213, p-value = 0.035) have a positive and significant influence on goal-depended creation modes. Accordingly, we accepted H3-6, H3-9, H3-10, H3-11, and H3-12.
Table 4-23: Effect of Organizational Culture on Knowledge Creation Modes KCM Facet Construct/Model
G-DR G-FR G-DE
MC 0.184 0.139 0.186*
AC -0.057 0.182+ 0.166*
HC -0.049 0.050 0.247*
OC
CC 0.133 0.196 0.213*
Adjusted R2 0.022 0.222 0.495 F-value 1.836 11.838*** 38.191***
N 153 153 153
1. + p < 0.1; *p < 0.05; ***p < 0.001
2. OC= Organizational Culture; MC= Market Culture; AC= Adhocracy Culture; HC= Hierarchy Culture; CC= Clan Culture; KCM= Knowledge Creation Modes; G-DE = Goal-Depended Creation Mode; G-DR = Goal-Driven Creation Mode; G-FR= Goal-Free Creation Mode
4.5 A Summary of Hypotheses Testing
The results of hypothesis test were listed in Table 4-24 in a concise manner. It was found that there were 33 hypotheses that were examined, and of which 18 hypotheses were supported.
Table 4-24: A Summary of Hypotheses Testing
H 1-1: Market culture has significant influence on product creation performance. Not support H 1-2: Adhocracy has significant influence on product creation performance. Support H 1-3 Hierarchy culture has significant influence on product creation performance. Not support H 1-4: Clan culture has significant influence on product creation performance. Not support H 1-5: Market culture has significant influence on manufacturing creation performance. Support H 1-6: Adhocracy has significant influence on manufacturing creation performance. Support H 1-7: Hierarchy culture has significant influence on manufacturing creation performance. Not support H 1-8: Clan culture has significant influence on manufacturing creation performance. Not support H 1-9: Market culture has significant influence on management creation performance. Not support H 1-10: Adhocracy has significant influence on management creation performance. Not support H 1-11: Hierarchy culture has significant influence on management creation performance. Support H 1-12: Clan culture has significant influence on management creation performance. Support H 2-1: Goal-driven creation mode has significant influence on product creation performance. Not support H 2-2: Goal-free creation mode has significant influence on product creation performance. Support H 2-3 Goal-driven creation mode has significant influence on manufacturing creation performance. Support H 2-4: Goal-free creation mode has significant influence on manufacturing creation performance. Support H 2-5: Goal-driven creation mode has significant influence on management creation performance. Support H 2-6: Goal-free creation mode has significant influence on management creation performance. Support H 2-7: Goal-depended creation mode has significant influence on product creation performance. Support H 2-8: Goal-depended creation mode has significant influence on manufacturing creation Support H 2-9: Goal-depended creation mode has significant influence on management creation performance. Support H 3-1: Market culture has significant influence on goal-driven creation mode. Not support H 3-2: Adhocracy culture has significant influence on goal-driven creation mode. Not support H 3-3: Hierarchy culture has significant influence on goal-driven creation mode. Not support H 3-4: Clan culture has significant influence on goal-driven creation mode. Not support H 3-5: Market culture has significant influence on goal-free creation mode. Not support H 3-6: Adhocracy culture has significant influence on goal-free creation mode. Support H 3-7: Hierarchy culture has significant influence on goal-free creation mode. Not support H 3-8: Clan culture has significant influence on goal-free creation mode. Not support H 3-9: Market culture has significant influence on goal-depended creation mode. Support H 3-10: Adhocracy culture has significant influence on goal-depended creation mode. Support H 3-11: Hierarchy culture has significant influence on goal-depended creation mode. Support H 3-12: Clan culture has significant influence on goal-depended creation mode. Support
4.6 Discussion and Implications
4.6.1 Organizational Culture to Creation Performance
In Table 4-21, the organizational culture is seen to be a significant predictor of creation performance. This result is consistent with those by Damanpour (1991), Deshpandé et al. (1993), Syrett and Lammiman (1997), Tushman and O’Reilly (1997), Chandler et al. (2000), and Martins and Terblanche (2003). This implies that companies in the manufacturing industry should be aware of the characteristics of organizational culture to improve their creation performance. Among culture characteristics, culture with characteristic of adhocracy has a positive impact on product creation performance (Lock and Kirkpatrick, 1995; Hauser, 1998; Mauriel et al., 2000; Martins and Terblanche, 2003). Moreover, market culture and adhocracy culture shows a positive and significant impact on creation performance in manufacturing. The result of impact of adhocracy culture on creation performance is consistent with Lock and Kirkpatrick (1995), Hauser (1998), Mauriel et al. (2000), Ogboma and Harris (2000) and Martins and Terblanche (2003). This indicates that an organizational culture with characteristics of market competition is more likely to emphasize goal achievement with respect to creation in manufacturing. In addition, organizational culture with characteristics of creation is likely willing to accentuate the entrepreneurship and likely willing to be a risk taker, such as the frequent adjustment of the internal and external requirements, the improvement of manufacturing processes and technologies, the manufacturing process reengineering, and the capability of order delivery.
In addition, organizational culture with characteristics of both hierarchy and clan is
more likely to positively enhancing the management creation performance. The result of the influence of clan culture on management creation performance is consistent with Angle (1989), Ogboma and Harris (2000), and Martins and Terblanche (2003).
However, Hauser (1998) argued that organizational culture with characteristics of clan would have negative influence on creation performance. Ogboma and Harris (2000) also proposed that hierarchy culture has negative influence on organizational performance. However, the participants selected by Ogboma’s and Harris’s research are from companies in United Kingdom. With other variables fixed, according to the standardized coefficient, it can be seen that clan (β = 0.389) have more contribution than hierarchy (β = 0.236). This implies that creation performance in management can be improved with an emphasis on organizational cohesion and loyalty, decision participation, and teamwork. The research suggestion is therefore that a company implementing a formal management process regulation and policy will stimulate employees to come up with creative ideas, and thereafter to improve management performance.
4.6.2 Knowledge Creation Modes to Creation Performance
First, although it is found from the SEM, the knowledge creation modes have not significantly influence on the creation performance. However, Deshpandé et al. (1993) proposed that the more knowledge creation modes adopt, the better creation performance is. Therefore, in order to explore more information between knowledge creation modes and creation performance, this research conducted the investigation of about the effects between sub-factors of knowledge creation modes and sub-factor of creation performance. In Table 4-22 it is found that both G-FR and G-DE positively and significantly influence the creation performance in product. The result of the
impact of G-FR on CP is consistent with Tesluk (1997) and Ayres (1999). However, G-DR does not show the same effect. With other variables fixed, the standardized coefficient of G-FR creation mode（β = 0.387）is larger than that of G-DE (β = 0.164) with respect to the contribution of product creation performance. This implies that in a creative company management with both G-FR and G-DE creation modes are more likely willing to frame the strategic decision (or goals) and then let their employees freely look for creative ways to reach the goals. By this, new products are more likely to meet the organizational strategy and characteristics of the creative products can be more diversified. Moreover, in the process of new product creation the goals can be slightly adjusted to relevantly meet the realistic requirements without putting the original goal(s) aside.
Second, with regards to the impact of knowledge creation modes it is found that, G-DR creation mode, G-FR creation mode, and G-DE creation mode have all significantly influence on the creation performance in manufacturing. Particularly, G-FR and G-DE have a positive influence while G-DR has negative. Based on the standardized coefficient, it is found that the G-FR presents the most contribution (β = 0.427) in affecting creation performance, and G-DE shows the middle (β = 0.290), and G-DR the last (β = -0.173). This implies that creation mode using goal-free is most likely to support creative idea generation in creation performance in manufacturing. In other words, normally G-FR creation mode gives employees more thinking space within a specific manufacturing technology or a specific manufacturing process to increase its performance. Besides, within a specific manufacturing requirement (or a goal) the G-DE creation mode may firstly set up a fundamental goal to frame the employees’ thinking space, and then shift to the G-FR
improvement. On the other hand, however, in adopting the G-DR creation mode, it can be seen from the obtained result that to highly limit thinking space may greatly weaken the creation performance in manufacturing.
Finally, three knowledge creation modes are all significantly related to the creation performance in management. However, similar to the creation performance in manufacturing, G-FR and G-DE has a positive impact while G-DR has negative.
Again, based on the standardized coefficient, it is found that the G-DE presents the most contribution (β = 0.505) in affecting creation performance, and G-FR shows the middle (β = 0.279) while G.-DR the last (β = -0.135). This implies that creation mode using goal-depended is most likely to support creative idea generation in creation performance in management. The reason is that a management strategy normally is made by top decision makers. Within a defined strategy, G-DE allows a free space to generate creative ideas for its implementation. On the other hand, G-DR creation mode does not have more rooms in raising ideas in creation of management. In consequence, adopting G-DR may lessen the performance of new management techniques.
4.6.3 Organizational Culture to Knowledge Creation Modes
Culture is in general regarded as a common value in things or a familiar behavior that society members inherit and may explicitly shape or control members in thinking and behavior (Tushmand and O’Reilly, 1997). It is found from Table 4-23 that organizational culture significantly influences knowledge creation modes (Angle, 1989; Nonaka, 1994; James and Roffe, 2000). This indicates that organizational culture is more likely to strengthen the ways (both thinking and behavior) that
employees perform their creation activities. First, it can be seen that organizational culture with four different characteristics (market, adhocracy, hierarchy, and clan) is positively and significantly related to G-DE creation mode. This implies that G-DE creation mode is significantly influenced by any characteristic of organizational culture. In other words, G-DE creation mode is suitable for any culture characteristics.
Therefore, it is our suggestion that manufacturing company may adopt G-DE creation mode while changing culture characteristic is not possible.
Second, for the impact of organizational culture on goal-driven creation mode, it is found insignificant from Table 4-23. This implies that no organizational culture characteristics can shape the mode of knowledge creation. In other words, a company who adopts a G-DR creation mode does not have to worry about what culture characteristic it has. Finally, for the impact on G-FR creation mode, there is no culture characteristic that shows significant. This is similar to the G-DR that a company can freely adopts the G-FR creation mode. In other words, the culture characteristics have no influence on the willingness of knowledge creation modes. However, if setting α to be 0.1, it is found that the culture with a characteristic of adhocracy has a positive and significant impact on G-FR creation mode. This suggests that organizational culture with a characteristic of creativeness leads significantly to a free environment of creation activities.
4.7 Differences of Sample Characteristics
In addition to the results, discussion, and implications described above, this research employed one-way Analysis of Variance (ANOVA) to further look at how different
characteristics of sample affect the research results. However, since the sample size is not large enough this research only considers three parts. One is based on the company age, one is based on the number of employee, and another is capital. Details are described below.
4.7.1 Company Age
First, according to the basic data, company age was divided into three levels: 1) below or equal to 10 years, 2) between 11 and 25 years, and 3) more than or equal to 26. The main objective is to examine the differences among these three categories with respect to the variables (organizational culture, knowledge creation modes, and creation performance). The hypothesis is defined to be that there is no difference for the mean of three categories. The results are listed in Table 4-25. It is found that there is no difference among these categories. This implies that research results described above are likely the same among three categories.
Table 4-25: ANOVA among Company Age Mean Composites ≦10 years
(N=49)
11 to 25 years (N=44)
≧26 years (N=60)
F-value p-value Decision
Organizational
Culture 5.088 5.307 5.328 1.430 0.243 Support
Knowledge
Creation Modes 4.737 4.763 4.771 0.035 0.965 Support
Creation
Performance 5.185 5.300 5.476 1.686 0.189 Support
4.7.2 The Number of Employees
Second, the number of employees is grouped into two categories. One is less than or equal to 1000 and another is the remainder. The hypothesis is defined to be that there is no difference between two categories of the number of employees with respect to organizational culture, knowledge creation modes, and creation performance. Results are given in Table 4-26. It is found that there is no difference between these categories.
This implies that research results described above are likely the same between different categories of the number of employees.
Table 4-26: ANOVA among Number of Employees Mean
Composites ≦1000 persons (N=89)
＞1000 persons (N=64)
F-value p-value Decision
Organizational
Culture 5.169 5.331 1.580 0.211 Support
Knowledge
Creation Modes 4.739 4.740 0.129 0.720 Support Creation
Performance 5.327 5.299 0.051 0.822 Support
4.7.3 Capital
Finally, the company capital is divided into two groups. One is less than or equal to 1 billion (denoted by G1) and another is greater than 1 billion (denoted by G2). The hypothesis therefore is defined to be that there is no difference between two groups of company capital with respect to organizational culture, knowledge creation modes, and creation performance. Results are given in Table 4-27. It is found that there is
difference between two categories for two variables (organizational culture and knowledge creation modes), and the mean of G2 is greater than that of G1. This implies that participants in G2 are more likely to grade higher for both knowledge culture and knowledge creation modes while answering the questionnaire. This implies that companies with higher capital are more likely to identify themselves a culture and creation mode with a particular characteristic. Moreover, a higher capital somehow implies a bigger company size. Therefore, a bigger company is more likely to confidently describe their culture characteristics and creation modes. In addition, it is found that there is no significant difference in creation performance for G1 and G2.
Statistically, this indicates that the capital can not be used to differentiate from creation performance.
Table 4-27 ANOVA among Company Capital Mean
Composites ≦1 billion NTD (N=68)
＞1 billion NTD (N=85)
F-value p-value Decision
Organizational
Culture 5.085 5.374 5.127 0.025* Not
Support Knowledge
Creation Modes 4.622 4.867 5.061 0.026* Not
Support Creation
Performance 5.189 5.413 3.257 0.073 Support
*p < 0.05
Chapter 5 Conclusions
From data analysis results in Chapter 4, Chapter 5 concludes the thesis. Limitations with respect to this study are addressed, and suggestions and future research focuses are delineated.
5.1 Research Conclusions
According to the data analysis result described in Chapter 4, the relations among organizational culture, knowledge creation modes, and creation performance were statistically disclosed. There are 18 hypotheses supported but 15 hypotheses unsupported. Because this model is a pilot study for knowledge creation modes, this result so far is used as an important reference. More studies may be needed to derive more substantial results. Conclusion can be made as follows.
5.1.1 Organizational Culture
Based on the results from structural equation modeling, organizational culture has a significant influence on creation performance (Damanpour, 1991; Deshpandé et al., 1993; Syrett and Lammiman, 1997; Tushman and O’Reilly, 1997; Chandler et al., 2000; Martins and Terblanche, 2003). The organizational culture has four different characteristics: competition, creation, hierarchy, and clan. The research results showed that organizational culture with characteristics of competition (market culture) would directly improve the creation performance in manufacturing creation performance
performance in manufacturing by adopting a goal-depended creation mode. Also, the culture with adhocracy can influence creation performance both in product and in manufacturing (Lock and Kirkpatrick, 1995; Hauser, 1998; Mauriel et al., 2000;
Ogboma and Harris, 2000; Martins and Terblanche, 2003). This characteristic can also make stronger via a goal-depended or a goal-free creation mode. The organizational culture with characteristics of hierarchy and clan can directly enhance the creation performance in management (Angle, 1989; Ogboma and Harris, 2000; Martins and Terblanche, 2003). Besides, the creation performance can be enhanced by the adoption of a goal-depended creation mode. By this, it is found that different culture characteristic would influence the creation performance in different areas (e.g. product, manufacturing, management). Particularly, any organizational culture characteristic can support the improvement of creation performance by using the goal-depended creation mode. In addition, the goal-depended creation mode is a supportive factor for an organization culture with the creative characteristic. Therefore, it is the research suggestion that companies who need to improve creation performance in areas should consider the culture they behave and the creation mode they adopt.
5.1.2 Knowledge Creation Modes
The research results for the research model disclosed that knowledge creation modes in whole is influenced by organizational culture (Angle, 1989; Nonaka, 1994; James and Roffe, 2000), but shows no significant impact on creation performance. When goes into their sub-factors to further explore the possible reasons. It is found from the results that goal-driven creation mode has negative and significant impact on creation performance in both manufacturing and management. This situation weakens the positive influence that goal-depended and goal-free creation mode contributes to the
creation performance. However, the value of coefficient (β = 0.308) shows the knowledge creation modes somehow can influence the creation performance.
Furthermore, the goal-depended creation mode is highly linked to any culture characteristics. This implies that particular culture effect may occur when companies in manufacturing industry take on and follow the goal-depended creation mode. It is the research suggestion that goal-depended creation mode may be a comparatively relevant way that a company can take on. For the goal-driven creation mode, however, the research finding indicates that there is no connection to the culture characteristic.
This implies that goal-driven creation mode may be adopted without concern of the culture characteristics. In addition, as mentioned before, there is a negative effect of knowledge-driven creation mode on creation performance in both manufacturing and management. It is our suggestion that it is not a good decision to adopt goal-driven mode if creation in manufacturing and management is concerned. Moreover, adopting goal-free creation mode in the organizational culture with the characteristic of adhocracy would improve the creation performance. In consequence, companies in manufacturing industry who would like to enhance the creation performance should consider the culture characteristic as well as the creation mode that they take on.
5.2 Research Contributions
1. Academic aspect
(1) Introduce and test knowledge creation modes in the research model:
Scriven (1977) first introduces the concept of creation mode. This concept has been depicted extensively in the literature. However, most of them are in the domain of
instructional projector program (e.g Evers, 1980). Few of them are in the evaluation of training performance (e.g. James and Roffe, 2000). In addition, despite of the evolvement of this concept in different areas, this study as a pilot research conducted an empirical study in industry to stimulate its use in knowledge creation domain as well as organizational behavior.
(2) Develop the measurement for knowledge creation modes:
Knowledge creation modes in industry are not new, but the measure is still in its infant, even in other domains. This research first goes deeply into the operational definition from the literature, and then develops its measure items accordingly.
These items may be a reference that continuous research in the same or similar domain can use.
2. Practical aspect
This study empirically investigates the causal relationships among knowledge creation modes, organizational culture, and creation performance. Research findings indicate that the characteristics of organizational culture do have significant impact on creation
This study empirically investigates the causal relationships among knowledge creation modes, organizational culture, and creation performance. Research findings indicate that the characteristics of organizational culture do have significant impact on creation