Impact of Organizational Climate on Knowledge-Sharing Behavior: An Empirical Study in Minimarkets

This empirical research aims to identify the relationship between organizational climate and knowledge sharing behavior. Also, this study uses knowledge sharing intentions as a mediator between these two variables. The survey was conducted incidentally on 403 minimarket employees in eight districts/cities in West Java. Factor analysis was employed to purify research variables. To evaluate the research model, multiple regression analysis was used. The results of the factor analysis on organizational climate resulted in two factors, fairness and affiliation. Fairness and affiliation constructs have a positive and significant effect on knowledge sharing intention. Knowledge sharing intention also has a positive and significant effect on knowledge sharing behavior.


INTRODUCTION
The minimarket business is growing rapidly in Indonesia (Hikmawatia & Nuryakina, 2017;Ginting, 2018;Hadyan, 2019). This business grew by 12% in 2019 (Hadyan, 2019;Pryanka, 2019). Minimarkets are the people's choice because of their close location and provide daily necessities (Hadyan, 2019). However, when there was a trade war between China and the United States in 2020, products from these two countries flooded Indonesia. The Covid-19 pandemic that has occurred since 2020 has triggered a business. Limitation of social activities and the obligation to maintain physical distance reduce the turnover of the minimarket business (Aria, 2020;Judith, 2021). As a result, retail business growth in 2020 is hampered (Pryanka, 2019;Kuncoro, 2020). Therefore, even though the Covid-19 vaccine has been discovered, retail business growth is estimated to be only 4-4.5% in 2021 (Susanto, 2020). The Covid 19 pandemic is perceived to cause uncertainty in 2021 (Business Wire, 2020; Sharma et al., 2020).
Moreover, in an uncertain business environment, the proactive action that minimarket managers can take is to innovate (Lv et al., 2018). Its motto, innovate or evaporate (Higgins, 1996). Innovation is known as a strategic process so that companies can adapt internally and externally (Henard & Szymanski, 2001;Grant, 2005), so that businesses can survive (Sorescu et al., 2003;Guan et al., 2009).
There are various drivers of innovation, one of which is knowledge (Lee, 2018;Castaneda & Cuellar, 2020). Knowledge is the main source of innovation (Wijekoon & Galahitiyawe, 2016). So, knowledge must be managed effectively and efficiently (Alavi & Leidner, 2001). Knowledge is useful in maintaining organizational existence (Drucker, 2000). There are various elements of knowledge management, but knowledge sharing is the most important element (Serenko & Bontis, 2016;Lee, 2018).
The problem is, knowledge resides in the individual (Nonaka & Konno, 1998;Smith, 2001). The organization is not the owner of these intellectual assets. Organizations cannot force employees to share knowledge with colleagues (Kelloway & Barling, 2000). So, organizations can only encourage or facilitate employees to voluntarily share knowledge (Gibbert & Krause, 2002).
Therefore, employees need to be motivated to share knowledge with their co-workers (Husted & Michailova, 2002;Wittenbaum et al., 2004). This can be done through the creation of contextual strength (Yoo & Torrey, 2002), in the form of organizational climate. Employees will interact intensively in a favorable organizational climate. When an employee has a problem, employees will interact with each other, actively participate in finding solutions (Hoegl et al, 2003). When information flows freely, employees will trust colleagues and management (Hinds & Pfeffer, 2003). Thus, the organizational climate has the potential to drive knowledge-sharing behavior (Razzaq et al., 2017;Andretto et al., 2019).
However, the behavior of employees to share knowledge is also determined by their intention (Fishbein & Ajzen, 1975). The intention to share knowledge is a direct precursor to knowledgesharing behavior (Ryu et al., 2003). In other words, the intention to share knowledge can be a mediator between organizational climate and knowledge-sharing behavior (Mafabi et al., 2017). Therefore, it is necessary to look for empirical evidence that can support the relationship between organizational climate and knowledge sharing behavior in the context of the minimarket business. To the best of the researcher's knowledge, this research did not yet exist.

HYPOTHESIS DEVELOPMENT
Organizational climate is the employee's perception of the company's social environment and policies (Patterson et al., 2004). Organizational climate is a reflection of what employees feel about organizational treatment (Harris, 2002). Schneider et al. (2011) define organizational climate as formal and informal perceptions of organizational policies, practices, procedures, and routines. Bock et al. (2005) used a facet-specific approach to identify dimensions of organizational climate that promote knowledge sharing. The results of the thematic analysis produce three dimensions of organizational climate, namely: fairness, innovativeness, and affiliation (Bock et al., 2005).
Fairness is an employee's perception of organizational practices that are reasonable, not arbitrary, and also not easy to change. Consequently, it creates trust among members of the organization to solve the "public goods dilemma" associated with knowledge-sharing problems (Bock et al., 2005). This construct is accepted as a driver that initiates knowledge sharing in organizations (Burgess, 2005). If employees believe they are evaluated and rewarded fairly, then they have a desire to contribute (Hislop, 2003). This can happen because fairness creates organizational citizenship and trust among organizational members (Bartol & Srivastava, 2002). So, when employees recognize a climate of fairness in the workplace, they tend to be more involved in various organizational activities, including knowledge-sharing activities (Bock et al., 2005). That is, a climate of fairness can motivate employee attitudes in building intentions to share knowledge. Thus, the following hypothesis can be derived. H1: Fairness climate has a positive effect on the intention to share knowledge.
Innovativeness is an iterative process to gain new opportunities by creating new findings (Garcia & Calantone, 2002). For organizations to always be innovative, organizational members must be encouraged to be open to the flow of information, oriented and focused on organizational learning, promote flexibility in doing routine work, support reasonable risk-taking, and reinforce entrepreneurial values (Roth, 2003;Bock et al., 2005). In an innovative climate, employees need to be prepared to anticipate changes, always looking for new and creative ideas (Hurley & Hult, 1998). Organizational members who work in an innovative climate will tend to share creative ideas throughout the organization (Kim & Lee, 1995). It is not surprising that an innovative climate is crucial for increasing knowledge sharing activities (Hsu & Wang, 2008). This allows the researcher to make the following hypothesis. H2: The climate of innovativeness has a positive effect on the intention to share knowledge.
Affiliation is a climate that will grow if there is a "sense of community" and pro-social behavior (Bock et al., 2005). Pro-social behavior encourages employees to voluntarily help colleagues they like or co-workers who are compatible with them (Chay et al., 2005). In other words, the sense of affiliation measures the sense of togetherness and openness among members of the organization. That feeling is built up by the warmth and attention the employee gets or when he or she requires that attention. Strong friendships in an organization encourage employees to work beyond their call of duty, to help each other in the organization (Bock & Kim, 2002). Employees with high affiliation tend to care about the feelings, thoughts, and views of their colleagues (Bock et al., 2005). As a result, those who have a high sense of affiliation will develop "strong bonds" with colleagues (such as friendship) and increase their social interactions (Cardador & Pratt, 2006). Furthermore, employees will be more attached to their co-workers and a "platform" will be created for them to share the knowledge they have. Conversely, employees with low levels of affiliation may not share knowledge even though the group they join has high collectivity (Ardichvili et al., 2006). Thus, a high level of affiliation is essential for knowledge sharing activities. It is not surprising that the high affiliation climate encourages employees to share knowledge (Trung & Thang, 2017). These arguments lead to the emergence of the following hypothesis. H3: The climate of affiliation has a positive effect on the intention to share knowledge. Ajzen (1975, 1981) stated that the precursor to behavior is intention. The intention is a person's location in the dimension of subjective probability that involves a relationship between himself and several actions (Fishbein & Ajzen, 1975). The intention is related to behavior. The relationship between intention and behavior is stated by Fishbein and Ajzen (1981) in Theory of Reasoned Action (TRA) and later modified in Theory of Planned Behavior (TPB; Ajzen, 1991). Various studies state that knowledge sharing behavior can be predicted by knowledge sharing intentions (Trung & Thang, 2017;Xue et al., 2012;Wang & Noe, 2010;Reychav & Weisberg, 2010;Chatzoglou & Vraimaki, 2009;Sihombing, 2009;Zhikun & Fungfai, 2009;Bock et al., 2005;Lin & Lee, 2004;Kolekofski & Heminger, 2002). In the context of knowledge management, a positive relationship between employees' desire to share knowledge and actual knowledge sharing behavior has been identified by researchers, this indicates predictive validity for employee behavior in organizations (Dawkins & Frass, 2005;Sutton, 2001;Sheppard et al., 1988). Therefore, the researcher can make the following hypothesis. H4: The intention to share knowledge has a positive influence on knowledge sharing behavior.

RESEARCH METHODS
This research uses a quantitative approach, categorized as explanatory (Singarimbun & Effendi, 2006;Creswell, 2008;2014), and is cross-sectional. Aims to explain the impact of organizational climate variables on knowledge-sharing behavior. In this case, the intention to share knowledge becomes the mediator of the two variables.
Minimarket employees in West Java are used as the unit of analysis. Researchers selected minimarket employees in four cities (Bandung, Bekasi, Cimahi, and Bogor), and four districts (Ciamis, Garut, Tasikmalaya, and Sumedang) in West Java. Sampling was carried out incidentally (Sugiyono, 2017). Minimarket permanent employees who are willing to fill out questionnaires are the respondents of this study. The distribution of questionnaires was carried out for one month, from January 15 to February 16, 2021, with a target of at least 50 respondents per city/regency. So, there are at least 400 respondents for eight cities/regencies. The number of respondents was determined based on the analytical tools used in this study, namely multivariate analysis in the form of multiple regressions (Hair et al., 2006;. Of the 900 questionnaires distributed, 403 were returned and filled in completely, or 45% of the questionnaires were eligible for processing. The survey was carried out through a structured and standardized close-ended questionnaire. The statement items for each construct were adapted from previous studies and measured by a five-point Likert scale; 1=strongly disagree and 5=strongly agree. Organizational climate has three dimensions, affiliation, innovativeness, and fairness (Bock et al., 2005). Affiliation consists of four statement items, for example, "The employees at this minimarket have a close friendship", have a reliability value, with the Cronbach alpha indicator of 0.8983. Innovativeness and fairness are each formed by three statement items; for example item for innovativeness "This minimarket encourages employees to improve the method or way of working", with a Cronbach alpha value of 0.8743. Examples of statement items for fairness "Bosses in this minimarket are objective, do not favoritism", with a Cronbach alpha of 0.8701. The intention to share knowledge variable consists of five statement items, adopted from Chennamaneni (2006); example item "If I have the opportunity, I will share my work experience with colleagues", has a Cronbach alpha value of 0.901. Knowledge sharing behavior is also adopted from Chennamaneni (2006), consisting of seven statement items, for example, item "I share my work experience with colleagues", with a Cronbach alpha of 0.92.
Researchers used factor analysis to purify the research instruments. Purification aims to evaluate the unidimensionality of items, whether they represent a single concept (Hattie, 1985;McDonald, 1981). Validity (Robinson et al., 1991) and reliability (Nunnally, 1979;Peter, 1979) were tested to ensure that the measuring instruments used in this study were valid and reliable. This study uses construct reliability (CR) and average variance extracted (AVE) to assess convergent validity (Hair et al., 2010). Furthermore, the research model was evaluated using multiple regression, and presented in the form of a structural model with path coefficients (Santoso, 2002;Hair et al., 2006;.

RESULTS AND DISCUSSION
This study involved 403 minimarket employees spread across eight cities/regencies in West Java. In general, the number of male and female employees is not much different, 57.6% male and 42.4% female. These employees come from seven minimarkets, dominated by Alfamart (41.2%) and Indomaret (44.7%). This is following the reality in the market, where the two minimarkets are the market leader. Furthermore, based on the age of minimarket employees, most of them are employees under 30 years old (95.3%). The rest are aged 31-40 years. The employees are generally high school graduates (54.1%) and vocational school (38.7%). Only 4.7% had a Diploma 1/2/3 education; 0.5% have a junior high school education, and 0.2% have a Master's degree. Finally, the working period of minimarket employees in this study was ≤1 year as much as 43.7%; 2-4 years as much as 46.9%; 5-6 years as much as 5.2%; and >6 years as much as 4.2%.
Researchers used the Kaiser-Mayer-Olkin (KMO) to measure the sampling adequacy test and the Bartlett sphericity test to evaluate whether the survey data were suitable for analysis (Hair et al., 2006;. Factor analysis is useful for determining construct validity, convergent, and discriminant. The KMO and Bartlett test results show that the data meet the basic requirements for analysis. KMO has a value of 0.896 and Bartlett's test is significant. Factor analysis was performed using the principal component analysis method, the rotation method was varimax with Kaiser Normalization. The result of the factor analysis presents four factors (from the five factors that want to be confirmed) with a total variance explained of 54.03%. The loading factor is only presented for values >0.60, useful for ensuring that it is practically significant (Hair et al., 2010). The results of the factor analysis, validity, and reliability analysis are presented in Table 1. One fairness item (coded "Fair1") joins three items used to measure innovativeness. However, because the loading factor of the four innovativeness items was less than 0.60, the four items were omitted (Hair et al., 2006). Thus, the construct of organizational climate has only two dimensions, namely: affiliation (the first factor, represented by items coded Affi1, Affi2, Affi3, and Affi4), and fairness (the second factor, represented by items coded Fair2 and Fair3). Innovativeness as an element of the organizational climate in minimarkets is dropping. This is quite interesting because several other studies (eg, Lin, 2007;Liao, 2006) stated that innovativeness is a consequence of knowledge sharing behavior. The third factor, the intention to share knowledge has a loading factor between 0.636-0.769, for the five statement items (Inten1, Inten2, Inten3, Inten4, Inten5). That is, nothing has changed from this construct. Different things happen in the fourth factor, knowledge sharing behavior. Initially, this construct consisted of seven statement items, but the results of the factor analysis left only three items (Behav2, Behav3, Behav5), four statement items were dropped because the loading factor was <0.60.
Thus, this study has identified four completely different factors or constructs, namely: fairness, affiliation, knowledge sharing intention, and knowledge sharing behavior.
Based on the results of the validity analysis, all statement items have a correlation value (item-total construct) between 0.404 (items coded Fair2 and Fair3) to 0.677 (items coded Inten3). So, the instrument used in this study is valid because it has a value of >0.30 (Sekaran, 2007;Robinson et al., 1991).
The researcher used the Cronbach alpha coefficient to evaluate four constructs (the innovativeness construct has been omitted). Only the fairness construct has a marginal Cronbach alpha value, while the constructs of affiliation, knowledge sharing intention, and knowledge sharing behavior have an acceptable value, >0.60 (Hair et al., 2006).
Information on loading factors from the results of the confirmatory factor analysis is then used to calculate construct reliability (CR) and average variance extracted (AVE) (Hair et al., 2010); useful for assessing convergent validity. The CR value ranges from 0.69-0.83; exceeding the recommended CR norm, namely: 0.70 or more. This indicates that there is internal consistency, that is, the statement items of the construct affiliation, fairness, knowledge sharing intention, and knowledge sharing behavior form a construct as expected. For AVE values, it ranges from 0.49-0.54; indicates a fairly good convergence, because the norm is ≥0.50 (Hair et al., 2010).
The mean, standard deviation, and bivariate correlation of the independent and dependent variables are calculated and reported in Table 2. A more detailed assessment shows that employees at minimarkets have a positive perception of the climate of affiliation and fairness. Knowledge sharing intention and knowledge sharing behavior are also positive, indicating that employees have the intention and behavior to share knowledge. If the mean value of the four constructs is classified into five classes (from "very low" class; between 1.00-1.80 to "very high" class; between 4.21-5.00), then all of the mean scores are classified into the "high" class; between 3.41-4.20. The standard deviation value is acceptable, which is a maximum of 20% of the mean value (Santoso, 2002). Following this, all constructs have a positive and significant correlation. A significant correlation indicates that the four constructs have interconnections (Ellonen et al., 2008). Thus, the criterion-related validity or predictive validity requirements are met. All independent constructs have a significant relationship with dependent constructs (Das et al., 2008). The independent constructs of affiliation and fairness have a relationship with the construct dependent on knowledge sharing intention. The construct of knowledge sharing intention (as an independent construct) is related to the dependent construct of knowledge sharing behavior. The root of the average variance extracted (√AVE) confirms the existence of discriminant validity (Fornell & Larcker, 1981) because its value is greater than the correlation value between constructs.
To evaluate the effect of organizational climate on knowledge sharing intentions and knowledge sharing behavior, researchers used multiple regression analysis. Visually, the results are reported in Figure 1

Ɛ2 Ɛ1
Note: Innovativeness as an element of the organizational climate in minimarkets is dropping in the factor analysis phase; because the loading factor value is <0.60. So, H 2 doesn't exist The constructs of fairness and affiliation have a positive and significant influence on the intention to share knowledge and can explain 15.9% of the variations in knowledge sharing intention. The affiliation has a greater influence (with a path coefficient of 0.291; t=5,596) on the intention to share knowledge compared to fairness (0.174; t=3.351). Knowledge sharing intention is proven to be a significant predictor of knowledge sharing behavior and can explain 19.4% of the variation in the knowledge sharing behavior construct.
Knowledge sharing intention acts as a mediator between fairness and affiliation with knowledge sharing behavior. This is because the regression results between fairness and affiliation with knowledge sharing behavior (directly, not through knowledge sharing intention) have a standard beta coefficient of 0.05 (t = 0.272) and -0.026 (t = 0.618) for p≤0.5 respectively. That is, the direct effect of the two constructs on knowledge sharing behavior is not significant. This is following the opinion of Baron and Kenny (1986) which states that the overall effect of the model without a mediator will be lower or become insignificant when compared to a model that integrates moderator variables. Because it is not significant, knowledge sharing intention is a full mediator between fairness and affiliation with knowledge sharing behavior (Baron & Kenny, 1986;Hair et al., 2010).

Impact Organizational Climate on Knowledge Sharing Intention
Conceptually, organizational climate consists of three dimensions, namely, fairness, innovativeness, and affiliation. However, minimarket employees have the perception that the organizational climate in their workplace is fairness and affiliation. This is in line with the study of Raharso and Tjahjawati (2014), which identified that innovativeness is a consequence of sharing knowledge in minimarkets, not a precursor. Thus, hypothesis 2 (H2) that innovativeness affects knowledge sharing intentions cannot be proven.
Fairness and affiliation have a positive and significant effect on the intention to share knowledge. That is, hypotheses 1 and 3 can be accepted, proven empirically. Thus, this study strengthens previous research that has succeeded in proving a relationship between organizational climate and knowledge sharing behavior (Yoo & Torrey, 2002;Bock et al., 2005;Erfan et al., 2013;Reyes & Zapata, 2014;Balozi et al., 2017). In this case, the organizational climate in the minimarket business is built by the dimensions of fairness and affiliation. Meanwhile, the climate of innovativeness is not a dimension of the climate prevailing in minimarkets.
This reflects the relationship of knowledge sharing among employees. In an atmosphere of fairness and positive affiliation, minimarket staff has high confidence in their colleagues, so that knowledge flows smoothly. This positive climate, of course, needs to be maintained, even improved to make it more positive. Because, if these efforts do not exist, then a positive work climate can become an unfavorable climate so that knowledge sharing activities are hampered. If a favorable climate is created, this situation will be very difficult to change (Ruggles, 1998).
Also, minimarket employees feel that they are treated fairly in the workplace, so they promote the organization's functions to run smoothly (Ya et al., 2003). Fairness creates trust among minimarket employees so that they can solve public goods dilemmas related to knowledgesharing problems (Bock et al., 2005). Knowledge is no longer a personal item but needs to be shared so that it is owned by other minimarket employees. Furthermore, minimarket employees trust each other with colleagues, thereby reducing transaction costs (Li et al., 2010) and promoting intellectual capital sharing (Nahapiet & Ghoshal, 1998). A high level of contact, from a social network perspective, will promote knowledge sharing among minimarket employees (Li et al., 2010).
These findings show that organizations that promote a favorable organizational climate will produce employees who consider information or knowledge as a common commodity, which is expressed in the form of knowledge-sharing behavior (Balozi et al., 2017).

Impact of Knowledge Sharing Intention on Knowledge Sharing Behavior
Furthermore, this research also strengthens the study of the influence of intention to share knowledge with knowledge sharing behavior, in the context of the minimarket business. The existence of a positive relationship between employee intention to share knowledge and actual knowledge sharing behavior is consistent with the results of previous research (Trung & Thang, 2017;Xue et al., 2012;Wang & Noe, 2010;Reychav & Weisberg, 2010;Chatzoglou & Vraimaki, 2009;Sihombing, 2009;Zhikun & Fungfai, 2009;Bock et al., 2005;Lin & Lee, 2004;Kolekofski & Heminger, 2002). This indicates predictive validity for employee behavior in organizations (Dawkins & Frass, 2005;Sutton, 2001;Sheppard et al., 1988). The role of intention to share knowledge as a "full mediator" between organizational climate and knowledge sharing behavior indicates the importance of the meaning of intention. That is, knowledge sharing intention fully predicts knowledge sharing behavior (Mafabi et al., 2017). This is parallel with Ajzen and Fishbein's (1980) statement which states that individual behavior is determined by the individual's intention to carry out the behavior. The intention is a cognitive representation of an individual's readiness to manifest behavioral intentions. The intention is considered as a close indicator of behavior (Ajzen, 1991). This means that this research strengthens the TRA and TPB studies which predict that behavioral intention is the most influential predictor of behavior (Sheppard et al., 1988), in the context of knowledge sharing in minimarkets.

CONCLUSION
The behavior of various knowledge is a social activity that cannot be forced, is voluntary. Consequently, organizations need to encourage this in a facilitative work context. One of them is to create a favorable organizational climate for knowledge sharing activities.
This research has identified two constructs that can promote knowledge sharing in the minimarket business; fairness and affiliation; innovativeness is not a climate in the minimarket business. Minimarket employees perceive these two constructs as factors that positively and significantly influence the intention to share knowledge. Furthermore, the intention to share knowledge has a significant effect on knowledge sharing behavior.