Subjective Norms or Psychological Empowerment? Moderation Effect of Power Distance on Knowledge Sharing

Abstract

This study examines the impacts of psychological empowerment and subjective norm on knowledge sharing in organizations, and the moderation effects aroused by power distance. Quantitative data from 567 valid questionnaires are collected by survey from a large company. The results demonstrate that subjective norm and psychological empowerment are positively associated with attitude toward knowledge sharing. Moreover, the findings further suggest that power distance undermines the influence of psychological empowerment on knowledge sharing but strengthens the effect of motivation to comply. Both theoretical and practical implications are discussed.

1. Introduction

Knowledge sharing has been considered a crucial element for the long-term survival of organizations [1,2]. An increasing number of organizations attempt to leverage the knowledge resources employees contribute for achieving desired organizational outcomes such as improving innovation performance, increasing productivity and maintaining competitive advantage [3]. Although it could be learned that knowledge sharing is especially significant for a company’s development, effective knowledge sharing is still a big challenge because knowledge is regarded as an important competitive source in most organizations [4].

A growing body of literature has explored the influential forces of knowledge sharing in maintaining sustainable competitiveness, where some studies demonstrate that organizations need to establish pro-sharing norms to promote employees’ contribution [1,5], while others argue that knowledge sharing within an organization is a kind of spontaneous situation. Thus, it is not always practical to force employees to share knowledge because the effective usage of knowledge management systems is an informal task that cannot be forced [6]. From the employees’ perspective, sharing knowledge is typically considered a double-edged sword, with the positive side being proactively sharing knowledge would be deemed as a “hero” for other employees, while the negative side is losing their expertise because the sharing process would cultivate new competitors [7,8].

Accordingly, current studies have not reached an agreement in this aspect. This leaves a large room for researchers to explore the potential impacts aroused by different organizational environments. Power distance is regarded as an essential factor in shaping employee preferences, which has garnered significant interest from numerous scholars [9,10]. Power distance refers to “the degree to an individual accepts unequal distribution of power” [10,11]. Understanding power distance is particularly crucial in organizational settings because power is the foundation of all relationships, which is also inherent in hierarchical organizations and subtly influences employees’ preferences and behaviors [12]. Experts in organization and social psychology have long demonstrated that, just as an individual’s perception of power distance can change, so do his/her autonomy, values and cognition [13,14,15,16]. However, it is still unclear whether and how employees’ values and motives shift along with the power distance.

With this direction, this study explores how power distance moderates the influence of different types of motivational factors on employees’ sharing knowledge. Drawing on the power distance theory and referring prior studies, this paper develops a theoretical model by integrating proactive and passive motivations and power distance. The contributions of this work are the following: First, employee knowledge sharing in organizations is always considered a complex issue for both researchers and managers [1,17]. Differentiating from prior research which focuses on either proactive motivation or passive motivation on knowledge sharing [6,18], this study organizes them together and understands how the differences would be, thus extending the current knowledge. Second, despite prior studies imply that an individual’s motivation may vary along the perception of power distance [13,14,15], it has not been meticulously examined whether and how power distance influences motivation varying along with the proactive to the passive. This paper explores how employees’ proactive and passive behavioral factors are influenced by their evaluations of organizations’ power distance, thus offering the potential opportunity to contribute to this field of knowledge. Third, this study also guides practitioners to design better strategies to encourage employees’ knowledge by understanding the relationship between knowledge sharing motivations and power distance.

2. Theoretical Background and Hypotheses

A growing number of studies have identified the influential forces that determine knowledge sharing in organizations [1,6,19,20], which is generally considered as proactive- and passive-oriented. Proactive motivation refers to the orientation that reflect an individual self-governing influence on proactively sharing knowledge [6,21]. Psychological empowerment captures the main aspects of proactive motivation. It is conceptualized as a formative construct manifested in a set of four dimensions: meaning (the fit between one’s perceptions and work), competence (one’s sense of self-efficacy), self-determination (autonomy over work or action), and impact (effect on work results) [6].

The consensus indicates that organizations need to establish pro-sharing norms to create a favorable atmosphere of knowledge collaboration [1,5]. Inherently, such pro-sharing norm acts as a mean to guide employees’ sharing behavior [22,23]. Specifically, attitudes or intentions formed through organization norms are part of external demands (e.g., the rules of conduct) which characterize a specific organizational setting. As a result, it is generally related to one’s behavior in an instrumental (or passive) way rather than a proactive way. This is consistent with prior research that the purpose of establishing organizational norms is to moderate employees’ behavior to meet the expectation of organizations [23,24]. A most commonly upheld passive-oriented motivation is the subjective norm, which has been supported by considerable empirical evidence [25]. Subjective norm refers to “the perceived normative pressure to perform or not perform an action” [26], which reflects the impacts of others’ expectations which are considered as important. It is largely contingent on a need for approval. To sum up, this study covers both types of motivations: proactive motivation-psychological empowerment and passive motivation-subjective norm.

2.1. Psychological Empowerment

According to the theory of psychological empowerment, psychological empowerment is a process of enhancing individuals’ conviction in self-efficacy, which reflects a proactive orientation to work [27]. Here, the proactive orientation mentioned reflects an orientation in which an individual expects to establish or sustain a harmonious work-role relationship with the organization. Therefore, influenced by psychological empowerment, employees may proactively undertake the responsibility of knowledge sharing, and are willing to form a positive attitude to help others solve work-related questions. A similar pattern was also obtained by prior research [6,28], where scholars found that psychological empowerment is positively associated with knowledge sharing. Thus, we propose that:

H1. 

There is a positive relationship between psychological empowerment and attitude toward knowledge sharing (henceforth, AttitoKS).

2.2. Subjective Norm

A series of social psychological theories indicate that subjective norm may powerfully shape one’s attitudes and intentions through perceiving social pressure. This implies that if employees realize that they are expected to take an action, they will have a more positive attitude [26,29]. In organizations, employees are often expected to share knowledge through knowledge management systems (henceforth, KMS) to create a favorable environment, thus achieving organizational goals [19]. When employees perceive pressure from organizational values or norms, they may take them as guidance and form a positive attitude. Prior studies suggest a positive relationship between subjective norm and AttitoKS [30]. Thus, we conjecture:

H2. 

There is a positive relationship between Subjective norm and AttitoKS.

2.3. Moderating Effects of Power Distance

With respect to the effects of psychological empowerment, it is conjectured that its impact on AttitoKS is contingent on power distance. Specifically, employees characterized by low power distance are more inclined to proactively make their own choices without necessarily considering their managers’ opinions because of the equal rights belief [31]. This expectation may facilitate employees to form a greater autonomy in issues of interest. Accordingly, employees in low power distance orientation environment tend to cultivate a sense of personal mastery or a “can do” attitude, so as to enhance their personal efficacy expectations. As mentioned earlier, psychological empowerment is a kind of proactive orientation that directly reflect one’s self-efficacy [27,32]. Thus, it could be inferred that the influence of psychological empowerment is expected to be stronger.

In contrast, employees with high power distance have a greater psychological dependence on organizational norms and tend to regard these norms as a basis for behavioral intentions [13,31]. As a result, their attitudes are more inclined to group rather than individual-determined. This claim is consistent with Matsumoto, Yoo and Nakagawa [9], which suggests that individuals with higher power distance orientation appear to be more inclined to obey organization norm without doubts. Consistent with this view, we conjecture that power distance will positively moderate the effects of subjective norm on AttitoKS. We therefore hypothesize:

H3. 

Power distance negatively moderates the influence of psychological empowerment on AttitoKS.

H4. 

Power distance positively moderates the influence of subjective norm on AttitoKS.

Figure 1 provides an overview of our research model. We next define each related construct and theoretically explain how the influential paths happen.

3. Research Methodology

3.1. Measurement

To empirically test the hypothesized relationships, we adapted the measurement scale designed from previous studies (as shown in

Table A1. Research constructs and measurements.

Construct	Item#	Measurement Items	Reference
Subjective Norm (formative construct)	6	Motivation to comply (MtoComply) 1. Generally speaking, I try to follow the manager’s policies and wishes. 2. Generally speaking, I accept and execute my boss’s decision, even if it differs from mine. 3. Generally speaking, I respect and put in practice the decisions of my colleagues. Normative beliefs on knowledge sharing (NOB) 1. My manager thinks that I should share my knowledge with my colleagues in the company. 2. My boss thinks that I should share my knowledge with my colleagues in the company. 3. My colleagues think that I should share my knowledge with them in the company.	[26]
Psychological Empowerment (formative construct)	12	Meaning dimension (EP_meaning) 1. The knowledge sharing I do is very important to me in KMSs. 2. My knowledge sharing activities are personally meaningful to me in KMSs. 3. The knowledge sharing I do is meaningful to me in KMSs. Competence dimension (EP_competence) 1. I am confident about my ability to share knowledge in KMSs. 2. I am self-assured about my capabilities to perform knowledge sharing in KMSs. 3. I have mastered the required skills for knowledge sharing in KMSs. Self-determination dimension (EP_self-determination) 1. I have a considerable autonomy in determining how to share my knowledge in KMSs. 2. I can decide on my own how to share my knowledge in KMSs. 3. I have a great opportunity for independence and freedom in how to share my knowledge in KMSs. Impact dimension (EP_impact) 1. My influence on what happens in KMSs is large. 2. I have a great deal of control over what happens in KMSs. 3. I have significant influence over what happens in KMSs.	[27]
Power Distance (reflective construct)	3	1. In my company, senior employees make most decisions without consulting the subordinates. 2. In my company, senior employees do not frequently ask the opinions of the subordinates. 3. In my company, senior employees rarely socialize with the subordinates.	[16]
Attitude toward Knowledge Sharing (reflective construct)	3	1. In my company, sharing knowledge with my colleagues through KMSs is good. 2. In my company, sharing knowledge with my colleagues through KMSs is harmful. 3. In my company, sharing knowledge with my colleagues through KMSs is a pleasant experience.	[26]
Knowledge Share Intention (reflective construct)	3	1. I intend to share knowledge in KMSs with my colleagues who need help/information. 2. I will try to take an active part in KMSs. 3. I will always make an effort to share knowledge in KMSs.	[26]

). All constructs were measured by using 7-point Likert-type scales, ranging from “strongly disagree” (1 point) to “strongly agree” (7 points). To ensure the consistency of the original meaning between Chinese and English version, each item was processed by a language instructor with a backward translation method and checked by five expert researchers in knowledge management fields [33]. In addition, the questionnaire was pre-piloted with 50 samples of representative employees to preliminarily test the reliability and validity.

3.2. Data Collection

A survey was distributed to the employees in a large company. The employees were asked to fill the survey when they were available in the focal week. The employees were informed with the objective to refine the knowledge management system in the company to realize better performance. Meanwhile, the employees were also told that their attendance is especially important for the company’s development in future. Finally, after the deleting of invalid samples that were not fully completed and inconsistent answers between reverse items and obverse items, 567 valid questionnaires were collected.

Table 1. Descriptive statistics of respondents.

Demographic Variable		Frequency	Percentage (%)	Cumulative (%)
Gender	Male	302	53.3	53.3
	Female	265	46.7	100.0
Age	20–30	327	57.7	57.7
	30–40	196	34.6	92.2
	40–50	40	7.1	99.3
	40 years or more	4	0.7	100.0
Education	High school or lower	10	1.8	1.8
	Bachelor’s degree	517	91.2	92.9
	Master’s degree or higher	40	7.1	100.0
Work experience (experience)	0–5 years	363	64.0	64.0
	5–10 years	145	25.6	89.6
	10–15 years	44	7.8	97.4
	15 years or more	15	2.6	100.0
Industry Sector (sector)	Service	305	53.8	53.8
	Manufacturing	262	46.2	100.0

demonstrates the demographic characteristics of the valid questionnaires.

3.3. Data Assessment

Before testing the theoretical model, it is necessary to evaluate the reliability and validity of all constructs. As shown in

Table 2. Reliability of Constructs.

Construct	Number of Questions	Cronbach’s Alpha	Composite Reliability
Empowerment-Meaning (EP_meaning)	3	0.943	0.869
Empowerment-Competence (EP_competence)	3	0.900	0.879
Empowerment-Self-determination (EP_determination)	3	0.870	0.860
Empowerment-Impact (EP_impact)	3	0.874	0.875
Normative belief on knowledge sharing (NoB)	3	0.934	0.891
Motivation to Comply (MtoComply)	3	0.810	0.872
Power distance (PowerD)	3	0.866	0.887
Attitude toward knowledge sharing (AttitoKS)	3	0.844	0.863
Knowledge sharing Intention (KnSI)	3	0.813	0.853

, the Cronbach’s Alpha (ranging from 0.810 to 0.943) and composite reliability (ranging from 0.853 to 0.891) values of all constructs are higher than the acceptable minimum value of 0.7, which indicate the reliability of measurements [34].

SPSS and SmartPLS software packages were employed to examine the validity [22,34]. Specifically, the convergent validity was checked with the principle of principal components analysis and varimax rotation [22]. Nine factors corresponding to the constructs appeared in exploratory factor analysis that was executed on all 29 items. Two items tapped onto other constructs and were omitted.

Table 3. Factor analysis (rotated factor matrix).

Item	Components
	1	2	3	4	5	6	7	8	9
EP_meaning1	0.285	0.121	0.805	−0.178	0.119	0.145	0.168	0.076	0.096
EP_meaning2	0.248	0.164	0.844	−0.197	0.140	0.182	0.162	0.083	0.046
EP_meaning3	0.236	0.158	0.840	−0.185	0.144	0.180	0.171	0.082	0.019
EP_competence1	0.827	0.111	0.305	−0.051	0.117	0.121	0.140	0.066	0.069
EP_competence2	0.879	0.133	0.209	−0.050	0.087	0.148	0.110	0.072	0.072
EP_competence3	0.818	0.095	0.145	0.016	0.139	0.174	0.133	0.077	0.027
EP_determination1	0.273	0.085	0.132	−0.167	0.184	0.764	0.146	0.092	0.052
EP_determination2	0.148	0.050	0.120	−0.106	0.166	0.865	0.112	0.014	0.032
EP_determination3	0.067	0.121	0.178	−0.083	0.180	0.829	0.068	0.107	0.054
EP_impact1	0.203	0.124	0.139	−0.120	0.779	0.212	0.060	0.110	0.038
EP_impact2	0.098	0.074	0.075	−0.127	0.854	0.194	0.031	0.148	0.039
EP_impact3	0.051	0.113	0.117	−0.112	0.874	0.110	0.034	0.111	0.015
MtoComply1	0.074	0.230	0.077	0.004	0.043	0.042	0.093	0.087	0.806
MtoComply2	0.069	0.046	0.006	0.115	0.007	−0.050	0.063	0.038	0.866
MtoComply3	−0.001	0.108	0.035	−0.054	0.030	0.122	0.077	0.029	0.825
NoB1	0.103	0.863	0.122	−0.106	0.130	0.092	0.160	0.181	0.160
NoB2	0.129	0.880	0.145	−0.152	0.144	0.065	0.149	0.156	0.148
NoB3	0.132	0.821	0.137	−0.082	0.074	0.117	0.194	0.186	0.170
PowerD1	0.035	−0.098	−0.138	0.851	−0.126	−0.090	−0.108	−0.083	0.045
PowerD2	−0.037	−0.082	−0.099	0.889	−0.116	−0.094	−0.051	−0.097	−0.011
PowerD3	−0.078	−0.099	−0.176	0.811	−0.090	−0.120	−0.062	−0.063	0.036
AttitoKS1	0.121	0.133	0.126	−0.029	0.018	0.089	0.825	0.141	0.092
AttitoKS2	0.100	0.133	0.103	−0.092	0.031	0.106	0.829	0.061	0.050
AttitoKS3	0.132	0.161	0.162	−0.106	0.070	0.091	0.814	0.108	0.119
KnSI1	0.065	0.161	0.024	−0.157	0.037	0.112	0.153	0.780	0.055
KnSI2	0.028	0.138	0.130	−0.085	0.130	0.128	0.148	0.830	0.059
KnSI3	0.099	0.135	0.037	−0.011	0.182	−0.043	0.011	0.826	0.048

Notes: KMO = 0.871, Bartlett’s Test of Sphericity: Chi-Square = 10,370.614, df = 351, Sig.= 0.000. Total Variance Explained = 80.912%.

showed that the rotated factor loadings. The results suggested the good convergent validity and discriminant validity. Meanwhile, to validate the discriminant validity, we further checked the correlations of related variables. As shown in

Table A2. Fornell–Larcker Criterion.

	1	2	3	4	5	6	7	8	9	10	11	12	13	14
1. Age	1
2. AttitoKS	−0.057	0.874
3. KnSI	0.051	0.318	0.852
4. EP_competence	0.107	0.355	0.237	0.913
5. EP_determination	−0.013	0.318	0.259	0.424	0.891
6. EP_impact	0.099	0.197	0.332	0.341	0.456	0.893
7. EP_meaning	0.012	0.418	0.279	0.567	0.457	0.38	0.947
8. Education	−0.085	−0.014	−0.033	−0.089	−0.041	−0.131	−0.062	1
9. MtoComply	0.005	0.239	0.175	0.174	0.148	0.117	0.169	−0.046	0.851
10. NoB	0.043	0.421	0.423	0.351	0.31	0.332	0.418	−0.106	0.357	0.94
11. PowerD	0.02	−0.234	−0.255	−0.159	−0.317	−0.316	−0.407	0.028	−0.001	−0.291	0.889
12. Sector	0.049	0.023	0.047	0.028	0.076	0.021	0.045	0.112	−0.084	−0.022	−0.045	1
13. Experience	0.701	−0.103	−0.001	0.049	−0.048	0.09	−0.09	−0.12	0.001	−0.02	0.096	−0.034	1
14. Gender	0.19	−0.064	0.024	0.085	−0.1	0.098	0.061	−0.109	0.039	0.009	0.055	−0.095	0.183	1

, the corresponding square roots of AVE values are much greater than the Pearson correlations in the measurement model [34]. In addition, as shown in

Table A3. Heterotrait–Monotrait Ratio (HTMT).

	1	2	3	4	5	6	7	8	9	10	11	12	13	14
1. Age
2. AttitoKS	0.063
3. KnSI	0.068	0.368
4. EP_competence	0.114	0.405	0.277
5. EP_determination	0.036	0.368	0.293	0.476
6. EP_impact	0.107	0.224	0.396	0.379	0.518
7. EP_meaning	0.02	0.466	0.31	0.614	0.502	0.415
8. Education	0.085	0.022	0.052	0.094	0.043	0.14	0.064
9. MtoComply	0.073	0.282	0.208	0.199	0.169	0.132	0.186	0.053
10. NoB	0.052	0.472	0.482	0.382	0.343	0.365	0.446	0.109	0.4
11. PowerD	0.021	0.271	0.293	0.18	0.364	0.363	0.45	0.034	0.097	0.323
12. Sector	0.049	0.033	0.059	0.03	0.082	0.039	0.046	0.112	0.092	0.022	0.048
13. Experience	0.701	0.111	0.084	0.053	0.052	0.097	0.092	0.12	0.081	0.025	0.102	0.034
14. Gender	0.19	0.07	0.052	0.089	0.108	0.107	0.063	0.109	0.093	0.018	0.061	0.095	0.183

, our results showed that the HTMT values of any pair of constructs are below 0.85, which demonstrates good discriminant validity [35].

For the second-order formative constructs of psychological empowerment and subjective norm, the significance of weights and multicollinearity assessment should be examined to validate the theoretical model [36]. As shown in

Table 4. Higher-order Construct Validation.

Second-Order	First-Order (Sub-Dimensions)	Weight	Significance	VIF
Psychological empowerment	Meaning dimension	0.382	p < 0.001	1.647
	Competence dimension	0.337	p < 0.001	1.567
	Self-determination dimension	0.309	p < 0.001	1.48
	Impact dimension	0.280	p < 0.001	1.337
Subjective norm	Motivation to comply	0.487	p < 0.001	1.145
	Normative belief on knowledge sharing	0.717	p < 0.001	1.145

, all sub-dimensions are statistically significant and the variance inflation factor (VIF) values are much lower than the acceptable maximum value of 5, indicating the formative constructs of psychological empowerment and subjective norm are valid.

3.4. Common Method Bias

Considering that all questionnaires were employees’ self-report, the common method bias might be a potential issue. To minimize the presence of potential threats, we followed the principles of Podsakoff et al. [37] in designing the online survey via ex-ante approach. The ex-post results of Harman’s single-factor test proposed by Podsakoff and Organ [38] suggested that the total variance explained by a single factor is 31.89%, which is much lower than the threshold. In addition, in line with Yang et al. [39], we used variance inflation factor (VIF) to further assess common method bias. As shown in Table 4, the VIF values ranged from 1.145 to 1.647, which are lower than 3.3 [39]. Hence, the common method bias does not seem to be a serious issue in our study.

4. Results

4.1. Main Findings

In order to test the structural model, model fit was checked [40]. We chose PLS because it provides bootstrap-based tests of exact model fit, by means of which a confirmatory composite analysis can be conducted. According to prior studies such as Benitez et al. [41], we selected several key indicators to examine the model fit. The results of

Table 5. Model fit indices.

Overall Fit of Estimated Model	Model Fit	Criterion	Reference
Standardized root mean square residual (SRMR)	0.049	<0.08	[40,41]
Normed fit index (NFI)	0.915	>0.90
dG (geodesic discrepancy)	0.161	dG < 95% bootstrap quantile (HI95 = 0.171)
dULS (unweighted least squares discrepancy)	0.151	dULS < 95% bootstrap quantile (HI95 = 0.187)

were computed with 5000 bootstrapped samples, which showed that the model fit is excellent and sufficient to proceed.

Furthermore, we used SmartPLS 3.0 software packages to test the research model [42,43]. Specifically, to determine the standardized estimates of the conceptual model, we empirically investigated the structural model using the PLS algorithm. In addition, we employed the bootstrapping method to estimate the significance and t-statistics of each hypothesis path within the structural model [44,45]. As shown in Figure 2 and

Table 6. Structural Model Results.

Hypothesis/Path	β	t-Statistic	p-Value	Conclusion
1. PsyEmpowerment → AttitoKS	0.286	6.098	0.000	Supported
2. Subjective norm → AttitoKS	0.250	5.659	0.000	Supported
3. PsyEmpowerment × PowerD → AttitoKS	−0.116	2.845	0.005	Supported
4. Subjective norm × PowerD → AttitoKS	0.039	0.965	0.335	Not supported
5. AttitoKS → KnSI	0.126	3.478	0.001	Supported
6. PowerD → AttitoKS	−0.032	0.693	0.489

, the standardized coefficients, significance levels and R² were illustrated for each path.

The results showed that psychological empowerment ($\beta =\begin{array}{l}0.286\hfill \end{array},p<0.001$) and subjective norm ($\beta =0.250,p<0.001$), operationalized as formative second-order constructs, exhibit significant influences on AttitoKS, supporting H1 and H2. In addition, the results further suggested that power distance has a negatively significant moderating effect on the relationship between psychological empowerment ($\beta =\begin{array}{c}-0.116\end{array},p<0.01$) and AttitoKS but has no significant moderating impact of subjective norm ($\beta =0.039,p=0.335$) on AttitoKS. Thus, Hypothesis 3 is supported but Hypothesis 4 is not supported. Furthermore, our results suggested a positive relationship between AttitoKS and knowledge sharing intention ($\beta =0.126,p<0.01$), which supports H5.

4.2. Mediating Effects Testing

Based on the principles for mediation tests proposed by Baron and Kenny [46] and Liden et al. [47], we took a four-step procedure for assessing the mediating role of the attitude toward knowledge sharing. They are: (1) test whether psychological empowerment and subjective norm are significant with knowledge sharing intention (Path A); (2) test whether attitude toward knowledge sharing is significant with knowledge sharing intention (Path B); (3) test whether psychological empowerment and subjective norm are significant with attitude toward knowledge sharing (Path C); and (4) add the direct effects of psychological empowerment and subjective norm on knowledge sharing intention and test the significance (Path D). As shown in

Table 7. Mediation effect test. ** p < 0.01, *** p < 0.001.

Path	Coef.	Std. Err.	Sign
Path A: PsyEmpowerment → KnSI	0.120	0.039	***
Path B: Subjective norm → KnSI	0.112	0.107	**
Path B: AttitoKS → KnSI	0.126	0.034	***
Path C: PsyEmpowerment → AttitoKS	0.279	0.040	***
Path C: Subjective norm → AttitoKS	0.249	0.045	***
Path D: PsyEmpowerment → AttitoKS	0.286	0.046	***
Path D: PsyEmpowerment → KnSI	0.104	0.039	***
Path D: Subjective norm → AttitoKS	0.25	0.044	***
Path D: Subjective norm → KnSI	0.096	0.045	**

, we conclude that attitude toward knowledge sharing plays a partial medium role in the relationship between psychological empowerment, subjective norm and knowledge sharing intention.

5. Discussions

This study explores whether and how psychological empowerment and subjective norm affect employees’ AttitoKS and the moderating effects aroused by power distance. The findings show that subjective norm and psychological empowerment significantly affect employees’ AttitoKS and knowledge sharing. This is consistent with previous studies which suggest that psychological empowerment is a crucial influential force in one’s decision about whether to share knowledge [6,28,48]. The findings imply that employees may proactively undertake the responsibility of knowledge sharing under the influence of psychological empowerment and are willing to form a positive AttitoKS. Meanwhile, as in line with prior works in examining the influence of norms [1,26,30] and social psychological theories [29,49]. This study also demonstrates the positive effect of norms on knowledge sharing, which further implies that subjective norm may powerfully shape ones’ attitude through perceiving social pressure. Thus, employees may share knowledge when others, such as bosses, colleagues and managers, take it as an essential task.

With respect to the moderation effects, as predicted, the findings reveal that power distance undermines the impact of psychological empowerment on AttitoKS. However, interestingly, there is no significant moderation effect being uncovered regarding the impacts of subjective norm on knowledge sharing behavior. This leads us to further deeper investigate the underlying reason regarding this. Our careful review of the extant literature shows that there are different conceptualizations in prior literature. Specifically, some studies conceptualize the construct as a reflective construct which is measured by overall perception on peer influence [1,19], while other studies typically conceptualize it as a second-order formative construct [26,29], where the scholars typically take it as a formation of different constructs, namely, the motivation to comply and normative belief on knowledge sharing. Conceptually, the motivation to comply and normative belief on knowledge sharing reflect different levels of subjective norms, where the normative belief on knowledge sharing is more related to the employee’s attitude regarding the knowledge sharing under the influences of other members, such as “the CEO thinks that I should share my knowledge with other members”, but the motivation to comply inclines to reflect the employee’s active attitude towards others’ decision, which could be measured such as “I respect and put in practice my colleague’s decision”.

As with different levels of conceptualization in the construct of subjective norms, we try to take the subjective norm as different dimensional reflections. This further leads us to consider whether these two dimensions generate different impacts on knowledge sharing. Accordingly, we conducted a post hoc analysis to further examine whether power distance had differential moderating effects on motivation to comply and normative belief on knowledge sharing. Figure 3 depicts the results of the post hoc analysis. Based on the post hoc analysis, we found that motivation to comply, normative belief on knowledge sharing and psychological empowerment are positively related to AttitoKS. In terms of moderating effect, power distance has a significant and positive moderating effect of motivation to comply on knowledge sharing. However, power distance has no significant moderating effect on the relationship between normative belief and knowledge sharing.

5.1. Theoretical Contributions

This study offers the following theoretical contributions. First, the findings broaden the knowledge sharing related literature. The employees’ knowledge sharing in organizations is always considered a complex issue, which attract extensive attention from both researchers and managers [1,17]. Despite prior research has revealed rich research implications, little has been uncovered about how employees’ knowledge sharing behavior varies along with their proactive and passive motivations. Differentiating from prior research which focuses on either proactive motivation or passive motivation on knowledge sharing [6,18], this study understands how the differences would be with an integrative way, thus extending the current knowledge.

Second, the present study enriches the theorization of motivation by exploring the role of power distance and nuanced the understanding of how multiple motives coexist in organizations. Previous research has implied that an individual’s motivation varies along the perception of power distance [13,14,15]. However, it has not been meticulously tested to demonstrate whether and how power distance influences motivation varying along with the proactive to the passive. The findings of this study fill this gap, thus extending the current understanding of this topic. The results suggested that the moderating effect of power distance on motives is contingent upon whether the motivation is proactive or passive. That is, the more proactive motivation, the more likely that power distance undermines the effect of motivation; however, the more passive motivation is, the more likely to strengthen the incentive effect.

Third, along with the argument of Petter, Straub and Rai [44] where the authors suggest that the failure to effectively valid the scientific results may be due to the fact that many studies often ignore the relationship between measurement items and constructs. The findings also remind researchers in future to more carefully study the formative vs. reflective constructs in measurement. That is, different conceptualizations and context-oriented measurements should be considered in academic research. As implied in this study, we may learn that the construct subjective norm could be differently conceptualized as formative vs. reflective, which may lead to more rich findings with this aspect [50,51,52].

5.2. Practical Implications

From a managerial point of view, this study provides an understanding of how to improve employees’ motivation to share knowledge in the context of Chinese enterprises. First, our findings show a positive relationship between psychological empowerment and AttitoKS. Thus, to form a perception of psychological empowerment, it is very important to incorporate a working environment that encourages participatory work and attaches importance to personal contribution into KMS [6]. Specifically, considering the high power-distance orientation of Chinese culture [53], the implementation of psychological empowerment involves changes to the traditional hierarchal structured organizations and doing so is not easy. Managers should create a supportive atmosphere and engage in confidence-building practices that inspire employees to take greater responsibility for their work. To this end, managers should articulate the overlap between organizational and individual goals to employees, and encourage them to be active in decision making. In addition, managers can develop employees’ empowerment in KMS as a way of motivational orientation for knowledge sharing. This development of KMS employee empowerment is inseparable from the joint efforts of the human resources department and the information system department.

Second, our findings show that subjective norm is a crucial influential force of knowledge sharing in organizations. Therefore, if organization managers expect a positive attitude of their employees toward knowledge sharing, they need to adopt serval practical strategies to promote the formation of subjective norms. On the one hand, influenced by Confucianism culture, Chinese employees strongly respect authority and prioritize group benefits [11,54]. Thus, senior managers should emphasize the importance of knowledge sharing in KMS and inform employees that sharing valuable knowledge is a form of contribution to the organization. On the other hand, managers should curb the impacts of losing their competitive advantage concerns of employees. They should thoughtfully design and execute their ongoing knowledge sharing strategies in order to make employees’ knowledge and experience become a part of organizational capital. For instance, they can adopt measures such as intellectual property protection and reward mechanism to eliminate the loss their expertise concerns of employees [1].

Finally, the improved power distance orientation might weaken employees’ proactive motivations. As a matter of fact, due to the high-power distance nature of Chinese society [53,55], it is difficult for managers to realize the unintended consequences of power distance orientation. Our findings show that power distance may shape one’s recognition and initiative in a systematic way. That is, it positively moderates the influence of motivation to comply on knowledge sharing, whereas it negatively moderates the proactive-based forces (e.g., psychological empowerment). If power distance orientation is higher, employees may have a greater psychological dependence on organization norms and tend to regard them as a basis for their behavioral intentions [13,31]. However, in cases where the power distance is low, the weight of psychological empowerment should be strengthened, and the effect of motivation to comply can be downplayed. Thus, appropriate guidance can help employees decide where to devote their efforts, reducing the lack of motivation.

5.3. Limitation and Research Directions

Like most previous studies, this work also has its limitations, which offer opportunities for further exploration. First, the research data in our study is a single source with cross-sectional, which may limit the generalizability of the findings. In line with the argument from previous studies such as [18,26] where the authors suggest that cross-sectional research design is not an ideal way to process the causality. We thus call for future research to explore the issue in relation to employee’s knowledge sharing by combining subjective data (reflected by employee’s psychological state) with objective data (reflected by employee’s actual behavior) with longitudinal data.

Second, the data was collected from a single research context may also restrict the generalizability of the results. For example, the differences may exist across different regions (such as Eastern and Western regions) and types of enterprises in terms of cultures, institutions and organization climates [56,57]. Accordingly, we further call for more research to extend the current findings by exploring and comparing different types of enterprises across various countries and regions.

6. Conclusions

This study attempts to improve current understandings of knowledge sharing by establishing a theoretical model integrating various types of motivations, power distance and knowledge sharing; to do so, this research proposes an integrative model to explain how employees’ knowledge sharing behavior varies along with their proactive and passive motivations. Overall, our findings suggest that power distance may shape one’s recognition and initiative in a systematic way, that is, it can compensate for the proactive motivation of psychological empowerment, while boosting the passive motivation of motivation to comply.