M.Sc. BA, specialization Organizational and Management Control
Supervisor: dr. B.J.W. Pennink
Second Assessor: mr. dr. W. Kaufmann
The Relation between Humanness and Knowledge-‐
Sharing within the Field of Accounting:
Empirical evidence from the comparative
examination of the States of former East and
former West Germany
by
NATASCHA K. PRÖSCHEL
University of Groningen
Faculty of Economics and Business
Abstract
The purpose of this study is to discover whether there is a positive causal relation between the cultural concept of humanness and the business concept of knowledge-‐sharing as well as the two constructs’ dimensions. As knowledge-‐sharing is a crucial element within the field of accounting, this study focuses on this specific context in order to propose a way to decrease any extant knowledge-‐sharing hostility. Centre of attraction are former East and former West Germany as two exemplary societies of distinct social value orientation, since it is expected that the latter would impact the aforementioned relation between humanness and knowledge-‐sharing. Data was empirically obtained via web-‐based questionnaires and subsequently analysed by various ANOVAs, Pearson product-‐moment correlations and linear regression analyses. The findings confirm the assumption that humanness can in fact provide a means of enhancing knowledge-‐sharing within the field of accounting. However, social value orientation is found to have no influence on the relation between humanness and knowledge-‐sharing. The paper thus attempts to draw attention to the beneficial use of humanness in order to establish a knowledge-‐sharing attitude within an organization and asks practitioners, i.e. managers and employees within the field of accounting, to actively incorporate humanness values in their everyday work.
Key words: humanness (Ubuntu); knowledge-‐sharing; accounting context; social value
orientation; former East vs. former West Germany
Supervisor: dr. B.J.W. Pennink
3
1
Introduction
Despite the increasing importance of knowledge-‐management within the field of accounting due to ever further developing information structures (Bhimani & Willcocks, 2014), there prevails vast resistance to knowledge-‐sharing (e.g. Bechina & Bommen, 2006; Chow et al., 2000; Husted & Michailova, 2002; Ragab & Arisha, 2013; Vera-‐Muñoz et al., 2006). The opposition is reinforced through organizations’ ignorance to appreciate such sharing acts (Lam & Lambermont-‐Ford, 2010; Ragab & Arisha, 2013). This constitutes a crucial problem especially in the accounting context as such fragmentation of knowledge can lead to failures in internal control or inferior financial decision-‐making (Lambert & Sponem, 2012), which is often harmful to the whole organization (Vera-‐Muñoz et al., 2006). Approximately 90 percent of any organization’s knowledge is said to be stored in its employees’ brains (Vera-‐Muñoz et al., 2006; Wah, 1999). However, only when knowledge is shared, it can gain in value because employees are able to benefit from each other’s wisdom. It is therefore of utmost importance to persuade accountants that rather knowledge-‐sharing, contrary to the mere possession of knowledge, constitutes power (Chong et al., 2011; Vera-‐Muñoz et al., 2006). Many challenges such as technological and behavioural barriers are yet to be overcome to reach an open knowledge-‐sharing environment (Chow et al., 2000). Managers from knowledge-‐intensive areas are thus keen to learn how to cope with the obstacles of knowledge-‐sharing (Foss et al., 2010). It is therefore necessary to uphold their endeavour and to avoid disappointments as in the past, when organizations recognized the significance of knowledge-‐management, but efforts to advance organizational knowledge have failed (Chase, 1997; De Long & Fahey, 2000).
Fulfilling Moilanen’s demand (2007) for a comparison of countries under the influence of the former collectivistic Soviet Union with countries of the individualistic Western world, the research compares the states of former East and former West Germany (hereafter referred to as East and West Germany) regarding their attitude toward humanness and knowledge-‐sharing. This has, to the author’s best knowledge, not yet been done in research.
The combination of East and West Germany was chosen for three main reasons. First, this research intends to investigate the influence of a major change in a political system such as the Wende in Germany on a society’s attitude and values. Both the communist German Democratic Republic (GDR), former East Germany, and the democratic Federal Republic of Germany (FRG), former West Germany, had made massive attempts from 1949-‐1990 to reshape their respective cultures. After 41 years of distinct radical regimes the two societies had indeed moved some distance apart (Inglehart, 1997). The Wende as major revolution in the two political systems gives this research reason to study the two societies after its reunification in 1990.
5
(e.g. Alesina & Fuchs-‐Schündeln, 2007; Friehe & Mechtel, 2014; Niemitz, 2009; van Hoorn & Maseland, 2010) confirm the still prevailing differences in values and preferences, which are also seen in reappearing terms such as wall in the head (e.g. Corbett, 2004) and Ostalgie (e.g. Zeitchik, 2003). Furthermore, the not ending media coverage in Germany and abroad indicates that the topic is still of great relevance (e.g. Meyer, 2014; Mayer & Theile, 2014). According to Alesina and Fuchs-‐Schündeln (2007) it will last between 20 and 40 years until dissimilarities between East and West Germany have almost fully vanished due to the combination of two reasons: the dying of the aged population and the actual shift in values.
Third, the comparative examination of East and West Germany is interesting from an accounting point of view, as the accounting traditions under the influence of the former Soviet Union, i.e. East Germany and of the West, i.e. West Germany differ due to distinct social and business conceptions (e.g. Bailey, 1995; Evans, 2004; Moilanen, 2008). Accounting under the influence of the former Soviet Union was not purposeful in the Western sense as it was disconnected from decision-‐making, planning and actual operations and was neutralized due to artificial prices. It acted merely as a tool for central government to lead (e.g. Bailey, 1995; Moilanen, 2008). Furthermore, many issues were handled informally for instance by relying on personal connections such as the social network blat (Hutchings & Michailova, 2004; Ledeneva, 1998; Michailova & Worm, 2003; Moilanen, 2008). The still existing network functions on the level of reciprocal connections and help in which the members pursue common interests and show a concern for the common well-‐being (Hutchings & Michailova, 2004; Ledeneva, 1998; Moilanen, 2008; Triandis, 1988). Contrary, the Western accounting model has increasingly moved towards decision-‐making, planning and actual operations (e.g. Granlund & Lukka, 1998; Moilanen, 2008; Scapens, 2005).
the accounting context. This study therefore aims to contribute to current research by providing insights into the knowledge-‐sharing behaviour within the accounting profession. This will be done in the specific case of Germany, while focusing on neglected areas such as micro level constructs and informal aspects (Foss et al., 2010).
Furthermore, this study attempts to fill a gap in current accounting research: studies examining knowledge-‐sharing so far have largely focused on knowledge-‐intensive, professional service, accountancy or consultancy firms (e.g. Chong et al., 2011; Ditillo, 2004, 2012; Morris & Empson, 1998; Vera-‐Muñoz et al., 2006; Weiss, 1999; Whitemore & Albers, 2006). As Wang and Noe (2010) asked to investigate how a certain culture is constituted in professional communities of practice, this research concentrates on the accounting profession itself, i.e. accounting employees in accounting departments of any organization including the private and the public sector. Accounting departments occupy a prominent position within organizations, as their work is generated in interplay between themselves, operational management and top-‐management. They are powerful towards top management through frequent interaction but operational management too reacts to their conduct (Mouritsen; 1996). Considering the management accounting function’s change from a traditional bean counter to a business partner one, accounting departments gain even further in importance (e.g. Lambert & Sponem, 2012). By turning information into action knowledge is created, which describes how the management accounting function contributes to the effectiveness of the organization (Bhimani & Roberts, 2004). Understandably, accountants are called “primary source of knowledge for most organizations” (Chong et al., 2011: 498). Since knowledge increasingly becomes the driving source for competitive advantage (e.g. Ditillo, 2004), management accounting will be progressively assessed according to its influence on knowledge-‐management (Bhimani & Roberts, 2004). This research therefore makes a first step to eliminate the current research gap and makes the results more useful for practitioners for any kind of organization.
7
The remainder of the research is structured as follows. The next section covers the relevant theoretical background regarding knowledge-‐sharing in general, knowledge-‐sharing within the accounting context, humanness and the social value orientation, i.e. the influence of East and West Germany. Based on this literature, the hypotheses will be stated ending in a conceptual model. Section three outlines the study’s methodology. Section four presents the research’s results analysed via various ANOVAs, Pearson product-‐moment correlations and linear regression analyses. Section five offers an explanation of the findings together with theoretical and managerial implications and states limitations along with suggestions
2
Theoretical Background
This section outlines relevant literature for a comprehensive understanding regarding humanness and knowledge-‐sharing to build up the relation with the social value orientation. Boom & Pennink’s study (2012) forms the basis for the present research’s framework. Based on the theoretical background, two hypotheses are formulated. The section concludes with the conceptual model developed for this research
.
2.1 The Concept of Knowledge-‐Sharing
2.1.1 Knowledge-‐sharing as competitive advantage. Today’s world is facing a transition towards a fast-‐moving knowledge-‐based economy which introduces new areas like knowledge management (e.g. Potecea et al., 2008). This research focuses on the concept of knowledge-‐sharing as the cornerstone of knowledge-‐management (e.g. Baxter & Chua, 1999; Chong et al., 2011; Vera-‐Muñoz et al., 2006). The success of knowledge-‐ management ultimately depends on knowledge-‐sharing (Wang & Noe, 2010) as only then individual employees’ knowledge is turned into organizational knowledge (Foss et al., 2010). Lin et al. (2009: 26) see knowledge-‐sharing as “social interaction culture, involving the exchange of employee knowledge, experiences, and skills through the whole department or organization”. The definition, implying that cultural norms influence employees’ willingness to share knowledge (Boom & Pennink, 2012), suits this research’s purpose with its focus on in-‐house knowledge-‐sharing attending to internal needs (Taylor & Murthy, 2009). Knowledge thus means using information while incorporating one’s expertise, interpretations, values and feelings (Boom & Pennink, 2012; Tan, 2000). Knowledge and information do not have to be distinguished when examining knowledge-‐ sharing (e.g. Bartol & Srivastava, 2002; Wang & Noe, 2010). Furthermore, there is no expected benefit in separating tacit and explicit knowledge (Bhimani & Willcocks, 2014). Moreover, the terms knowledge-‐transfer, knowledge-‐exchange and knowledge-‐sharing are often used interchangeably (Foss et al., 2010). However, when making use of the first two terms, one must only focus on the individual’s sharing act itself (Wang & Noe, 2010).
9
comprehension of intra-‐organizational processes could help accountants as experts in information compilation and distribution to enhance knowledge-‐sharing practices and systems (Chow et al., 2000). Research (e.g. Beck et al., 2014; Chow et al., 2000; Foss et al., 2010; Lin et al., 2009; Wang & Noe, 2010) has identified various mechanisms through which knowledge-‐sharing contributes to the organizational competitive advantage; nonetheless those are out of the scope of this paper.
Despite the importance of knowledge-‐sharing, great resistance prevails (Bechina & Bommen, 2006; Husted & Michailova, 2002; Willem, 2003) and knowledge is shared only little (Szulanski, 2000; Vera-‐Muñoz et al., 2006). Reasons range from wanting to protect one’s career chances, to spare time and resources or to prevent any unfortunate uncovering (Gammelgaard, 2004; Husted & Michailova, 2002). If organizations want to create an open knowledge-‐sharing environment they have to understand that this requires mutual trust as knowledge-‐sharing is of voluntary nature and involves a shift in power (Boom & Pennink, 2012; Lin et al., 2009).
The concept of knowledge-‐sharing consists of the following three dimensions which were identified as the most powerful factors of organizational knowledge-‐sharing in research: (1) corporate culture and leadership, (2) employee motivation, and (3) information technology (Boom & Pennink, 2012; Lin et al., 2009; Scholtens, 2011). (1) Corporate culture and leadership refers to a social-‐oriented organizational atmosphere, an orientation towards learning, management support and most importantly to interpersonal trust. (2) Employee motivation includes individual beliefs and values such as personal benefits and expected reputation. (3) Information technology has been incorporated as it becomes increasingly important in modern times by offering new opportunities for knowledge-‐sharing such as the sharing via knowledge-‐bases, communities of practices, the Internet or intranets (Boom & Pennink, 2012; Lin et al., 2009).
Especially if social cohesion surrounds a relationship, the tie positively influences knowledge-‐sharing (Reagans & McEvily, 2003).
2.1.2 Knowledge-‐sharing as crucial element in the field of accounting. Accountants
as knowledge-‐workers require a large amount of current quantitative information to uphold their expert status (Taylor et al., 2001). They require a combination of the following knowledge and skills: (1) a comprehensive technical knowledge to match information to the requirements of the respective situation; (2) a proper knowledge of other organizational functions such as marketing or production; (3) an array of social and interpersonal skills (Pierce & O’Dea, 2003). Furthermore, it is important to gather information via informal knowledge-‐bases from other business functions for instance via verbally communicated messages or organizational social networks (Vaivio & Kokko, 2006).
The management accounting function’s change from a traditional bean counter to a business partner one has been hotly debated among academics and professionals with field-‐ related literature supporting this transition (e.g. Bhimani & Willcocks, 2014; Bromwich & Bhimani, 1994; Granlund & Lukka, 1998; Kaplan, 1995; Lambert & Sponem, 2012; Mouritsen, 1996; Vaivio & Kokko, 2006). By forming organizational practice rather than informing it, accounting plays a major role in managerial work and should be considered from a broader managerial perspective, rather than an accounting one (Gerdin et al., 2014). This research therefore argues that based on the typology of Lambert and Sponem (2012), the trend goes toward a partner management accounting function where management accountants and operational managers engage in a so-‐called partnership. This particular management accounting style is assigned one major role, namely the assistance in operational decision-‐ making, which ascribes great importance to accounting departments in producing information (Mouritsen, 1996; Lambert & Sponem, 2012).
11
2014). Especially when geographical distance between departments increases, the importance of accounting information gains on weight, as it constitutes the major or even the sole carrier of knowledge (Roberts & Scapens, 1985). Various organizational activities can be categorized through accounting information, which then allows controlling them in an economically and managerially functional way (Bhimani & Roberts, 2004; Jones & Dugdale, 2001). Accounting thus contributes essentially to knowledge-‐management (Bhimani & Roberts, 2004), as it establishes a basis for the interconnection between the human, the knowledge and the financial resources. This suggests a new perspective for the accountant as resource controller who adds to the overall welfare and prosperity of the organization (Roberts, 2003).
2.2 Culture and the Concept of Humanness
Individual and organizational behaviour is embedded in the context of the prevailing national culture (Michailova & Hutchings, 2006). Many cultural researchers (e.g. Hofstede, 2001; Triandis, 1989) have therefore constructed paradigmatic characterizations regarding national values (Michailova & Hutchings, 2006). Culture is mirrored in values and practices and thus determines what organizations aim for and how they realize it (De Long & Fahay, 2000; Triandis, 1995). Furthermore, culture forms individuals’ actions (Chow et al., 2000; Erez & Earley, 1993; Hofstede, 2001; Michailova & Hutchings, 2006; Triandis, 1989) by telling them how to act under certain circumstances (Geertz, 1973; Möller & Svahn, 2004; Swidler, 1986; Triandis, 1995). Cultural patterns are hence deeply embedded and also shape individuals’ knowledge treatment and knowledge-‐sharing (Möller & Svahn, 2004). Still, the conception of culture remains broad and complex and makes research on its influence difficult (Geertz, 1973; Holden, 2002; Möller & Svahn, 2004; Swidler, 1986). To enhance the research’s feasibility, the study concentrates on the concept of ethnic culture as Möller and Svahn (2004) did, given the fact that a nation can consist of various ethnic cultures. The latter concept includes all of society’s different layers and its impact can be seen in people’s every day behaviour (Möller & Svahn, 2004; Triandis, 1995).
are said to be very helpful (Taylor & Murthy, 2009) which plays a central role in the concept of humanness.
Humanness is the English translation for the African philosophy Ubuntu (English, 2002; Lutz, 2009; Sigger et al., 2010). Its most widespread definition is umunutu ngumuntu ngabantu which means a person is only through other persons (Broodryk, 2006; Karsten & Illa, 2005; Mangaliso, 2001; Sigger et al., 2010). Humanness addresses humans’ responsibility towards each other and their interconnectedness (Nussbaum, 2003) and can hence be perceived as collective cooperation (Rwelamila et al., 1999). According to Ubuntu, a person’s status is decided on account of their interactions and relationships with other people rather than their power or money. Ubuntu has lately emerged as the so-‐called humanness management style, which focuses on people and stresses cooperating with and respecting each other (English, 2002). It stands out due to its focus on the whole community rather than single individuals and to people’s willingness to collaborate in view of a collective goal, to help each other and to share experiences (Lutz, 2009; Mangaliso, 2001). Humanness can thus be found in employees’ interaction with each other together with their behaviour in sharing for instance knowledge or ideas (Karsten & Illa, 2005).
Several researchers (e.g. Kamwangamalu, 1999; Lutz, 2009; Makhudu, 1993; Nussbaum, 2003; Prinsloo, 2000; Sigger et al., 2010; van der Colff, 2003) argue that humanness can readily be transferred to the Western context. Humanness values should therefore be called human values instead of African values (Lutz, 2009; van der Colff, 2003) as they can be found in various contexts around the world (Kamwangamalu, 1999). They manifest themselves to a varying degree for instance in political ideologies such as communism, socialism or capitalism or can be seen in Western thinking vocabulary such as sharing, dignity and trust or comrade within communism meaning brother within humanness (Prinsloo, 2000). Humanness values are gained through socialization processes such as linguistic or political contexts (Kamwangamalu, 1999) or are even general human values innate to every person (e.g. Makhudu, 1993; Prinsloo, 2000; Sigger et al., 2010; van der Colff, 2003). The prevalence of humanness in the Western context thus must not be overlooked as Western organizations might gain new insights from humanness as a management tool (Sigger et al., 2010).
13
Broodryk, 2006; Poovan et al., 2006; Sigger et al., 2010) have later on merged the closely connected dimensions respect and dignity and use four dimensions in their studies. All dimensions together constitute a collective value-‐system (Poovan et al., 2006). (1) Solidarity encompasses a collective achievement of demanding jobs and the responsibility towards society (Sigger et al., 2010). In an organizational context, solidarity contributes to greater team cohesion and commitment (Broodryk, 2006; Poovan et al., 2006; Sigger et al., 2010). (2) Survival means to engage in a collective attitude (Poovan et al., 2006) and share one’s resources to protect society (Boom & Pennink, 2012). In an organizational context, survival leads to greater effectiveness within the team (Mbigi, 1997; Sigger et al., 2010). (3) Compassion entails comprehending the difficulties of others and having the desire to help them (Poovan et al., 2006; Sigger et al., 2010). In an organizational context, compassion contributes to the development of a shared vision (Sigger et al., 2010). (4) Respect and dignity refer to the courteous treatment of each other (Boom & Pennink, 2012), which can enable effective performance in an organizational context (Mbigi & Maree, 1995; Poovan et al., 2006; Sigger et al., 2010). An important outcome of mutual respect is the increase in trust between individuals, which ultimately leads to the sharing of experiences and knowledge (Scholtens, 2011).
Sigger et al. (2010) built upon the humanness dimensions and developed a measurement tool regarding the presence of humanness which is according to the author’s best knowledge the only existing and validated measurement tool. The instrument can be used for research within organizational settings and enables the application within the Western context (Sigger et al., 2010). This research applies Sigger et al.’s (2010) measurement tool while adjusting it according to the recommendations by Scholtens (2011) and Boom & Pennink (2012) as well as partially combining it with the knowledge-‐sharing measurement tool adapted from Taylor and Murthy (2009).
2.3 The Relation between the Concept of Knowledge-‐Sharing and the Concept of
Humanness
a positive causal relation has been found between the degree in humanness and the degree in knowledge-‐sharing (Boom & Pennink, 2012; Fredriks, 2012; Scholtens, 2011), with the reason behind being that the idea of sharing is a very basic component to the concept of humanness (Makhudu, 1993; Prinsloo, 2000). Following Boom & Pennink’ study (2012) this research’s assumption is thus that an organization’s degree in humanness through the ethnic culture of its employees affects individuals’ willingness to share knowledge within the organization.
Several researchers (e.g. Dent, 1991; Hopper et al., 1987; Jones, 1992) understand accounting as being socially embedded within cultural structures including beliefs, values and knowledge: Those patterns finally determine individuals’ actions rather than accounting being a mere technical-‐rational activity. This reflects the Bourdieuan perspective in which accounting is seen as being situated in a fluid and subjective field where individuals’ attitudes merge with the social organizational structures and interactions (Bourdieu, 1990; Bourdieu & Wacquant, 1992). Furthermore, practitioners themselves regard accounting as an activity including objective as well as subjective and inter-‐subjective elements rather than purely objective ones (Jones, 1992). Hopper et al. (1987: 438), supported by further researchers (e.g. Armstrong, 1985; Burchell et al., 1980), describe accounting as “a fully social practice [that is] both the medium and the outcome of the politico-‐economic context in which accounting is embedded” (Jones, 1992). This perfectly explains the influencability of accounting, which serves as basis of this research (Gray, 1988).
Based on the preceding theoretical background the first hypothesis of this research is formulated as follows.
H1: Accounting employees scoring high in valuing humanness show greater willingness to share knowledge than those accounting employees who do not score high in valuing humanness.
It is expected that the collective nature of humanness will be mirrored in employees’ behaviour regarding a communal objective: Employees are anticipated to be willing to help, collaborate and share without any extrinsic reward (Lutz, 2009; Mangaliso, 2001).
2.4 The Social Value Orientation
15
facing situations of interdependence (Messick & McClintock, 1968). In this research the term refers to the distinctive ideological background of a certain society.
2.4.1 The individualism-‐collectivism dimension. The cultural dimension
individualism-‐collectivism is used most often when distinguishing how different communities analyse and process information and is hence appropriate when examining the concept of knowledge-‐sharing (e.g. Bhagat et al., 2002; Michailova & Hutchings, 2006; Möller & Svahn, 2004). Furthermore, the individualism-‐collectivism orientation influences how individuals behave at work (Erez & Earley, 1993; Hofstede, 1980; Triandis, 1989), which is of great significance in this research. Generally, the dimension refers to the relative importance that individuals of a certain community attribute to their own self-‐interests versus those of the whole group (Chow et al., 2000). The orientation thus relates to the degree of social connectedness between a member and their social community, which is especially in the world of work an essential distinguishing characteristic (Erez & Earley, 1993).
responsibility, in which flourishing relationships are placed subordinate to efficiency. The system therefore counters the concept of humanness (Mangaliso, 2001).
2.4.2 Former East and former West Germany as exemplary societies. To be able to
draw conclusions regarding the political influence of the two societies’ distinct regimes comparability had to be ensured: (1) East and West Germans’ values had not differed prior to the German separation; (2) the introduction of the two distinct regimes was not upon request of the citizens; and (3) during the period of separation the two societies were comparable in their most important dimensions (Friehe & Mechtel, 2014). This confirms that any dissimilarity in behaviour between the two societies might indeed be caused by the distinct political experiences (e.g. Alesina & Fuchs-‐Schündeln, 2007; Friehe & Mechtel, 2014; Heineck & Süssmuth, 2013; Rainer & Siedler, 2009). Any preference is thus to some degree traceable to preceding political regimes or to the political system still in power (Alesina & Fuchs-‐Schündeln, 2007; Friehe & Mechtel, 2014).
2.5 The Relation between the Concept of Humanness and the Social Value
Orientation
The individualism-‐collectivism dimension mirrors what kind of information individuals prefer and thus are more likely to process (Bhagat et al., 2002; Möller & Svahn, 2004). More specifically, collectivistic rather than individualistic societies tend to be more cooperative (e.g. Chow et al., 2000; Cox et al., 1991; Wagner, 1995) and show a greater degree in knowledge-‐sharing (e.g. Chow et al., 2000; Ma et al., 2014; Ragab & Arisha, 2013). The above confirms the supposition that humanness can be associated with a certain type collectivism (Prinsloo, 2000). Many parallels can also be seen between the cultural values of East Germany and the concept of humanness. The research therefore assumes collectivism, i.e. East Germany to have a rather high degree in humanness and individualism, i.e. West Germany to have a rather low degree in humanness.
Existing literature thus leads to the second hypothesis of this research.
H2: The more collectivistically-‐oriented the accounting employee’s degree of his/her social value orientation, the higher will be the association between humanness and knowledge-‐sharing.
17
cultural impact. It is thus expected that the social value orientation as the organizational social context will act as a moderator on the relation between humanness and knowledge-‐ sharing (Bourdieu, 1990; Bourdieu & Wacquant, 1992) in which it is anticipated that the relation will be strengthened in a collectivistic society, i.e. East Germany and weakened in an individualistic society, i.e. West Germany. From the literature outlined above it is expected that the distinction in the two societies’ social value orientation is rather big and that this affects the relation between humanness and knowledge-‐sharing.
2.6 Conceptual Model
To improve the study’s readability and to be able to make detailed descriptions of the relation between humanness and knowledge-‐sharing, data is analysed via three different levels: (1) level 1 – the relation between the two main concepts humanness and knowledge-‐ sharing; (2) level 2 – the relations between one main concept and the dimensions of the other concept; (3) level 3 – the relations between the dimensions of both concepts. The conceptual model in Figure 1 displays analysis level 1. It depicts the two main concepts humanness and knowledge-‐sharing and displays that humanness (independent variable) is expected to influence knowledge-‐sharing (dependent variable). Based on the theoretical background, this relation is anticipated to be causally positive. Furthermore, the relation is tested under the influence of the social value orientation (moderator) with the expectation of the relation being stronger in the case of a more collectivistic social value orientation.
FIGURE 1:
Level 1 – Relation between the main concepts
H1: Accounting employees scoring high in valuing humanness show greater willingness to share knowledge than those accounting employees who do not score high on valuing humanness.
3
Methodology
This section gives a methodological overview of the research process and discusses how data was collected, how participants were chosen, which scales and measures were utilized, how reliability and validity was assured and how data was analysed.
3.1 Data Collection and Participants
The present study got its bearings by previous studies (Boom & Pennink, 2012; Scholtens, 2011; Taylor & Murthy, 2009), which collected their data via questionnaires to measure humanness or knowledge-‐sharing respectively. A web-‐based questionnaire gained preference in the current context in which emails and online posts were used as contact method and the web as response method in order to collect primary, quantitative data. There were several reasons for applying a web-‐based questionnaire: (1) the target population could be contacted through various lists of e-‐mail addresses; (2) the questionnaire involved several questions of sensitive character which required confidentiality; (3) practical considerations such as the fast questionnaire distribution and data analysis or the higher data quality due to electronic database storing (Schonlau et al., 2002). The main reason was the possibility to approach a big number of potential participants to tackle the anticipated low response rate (Jansen et al., 2007).
As is often the case in research, time and budget constraints led the author to choose convenience sampling to approach a hard-‐to-‐reach population (Schonlau et al., 2002; Sue & Ritter, 2007; Thomas, 2004): the German accountancy world. This diminished adequate generalization of the research findings but not the relevance of the research purpose (Thomas, 2004). “It is better to have collected some data and gained some insight than to have collected no data and gained no information”, insomuch as decreased power in statistical inference and the possibly biased results are considered (Hill, 1998; Sue & Ritter, 2007: 34; Thomas, 2004).
19
As no formulas for the preferred sample size within nonprobability samples existed (Sue & Ritter, 2007), this research aimed for a sample of 100 to 200 participants to provide an appropriate frame for the data analysis (e.g. Bailey, 1994; Thomas, 2004), according to the general rule: “as large as it can be given your time frame” (Sue & Ritter, 2007: 34). 32 accounting associations were each twice approached via email with the request to forward the questionnaire to their members and partners. Furthermore, the questionnaire was posted twice in 22 accounting-‐related groups on the professional social software platform Xing1. Additionally, few private connections within the German accountancy world were
addressed. The questionnaire was fielded for three months and reached approximately 7100 (600 via accounting associations, 6500 via Xing) German accounting employees. Altogether 108 persons participated in the questionnaire; however eight participants were excluded from the data analysis, as they did not meet both inclusion criteria: (1) German nationality; and (2) employee within the field of accounting. The low participation rate might be explained by the proliferation of web-‐based questionnaires or the fact that no incentives, except for the possibility to receive the research’s results after the study’s finalization, were offered to the participants (Tourangeau et al., 2013).
Table 1 shows the distribution of the remaining 100 participants’ background. The greater amount of male participants contrary to female participants resembles results from literature (e.g. Bundesagentur für Arbeit, 2014; Hasselback & Carolfi, 1995; Taylor & Murthy, 2009). Furthermore, participants’ distribution in age and sector reflects the current German job market within the field (Bundesagentur für Arbeit, 2014). This suggests that the research is representative of German accounting employees and that non-‐respondents bias should prevail only to a limited degree.
TABLE 1:
Participants’ background
3.2 Scales and Measures
Coming from validated and accepted scales (Boom & Pennink, 2012; Taylor & Murthy, 2009), the study measured the relation between humanness and knowledge-‐ sharing and the influence of the social value orientation on an individual level of analysis through a web-‐based questionnaire consisting of 69 questions. The participants had never been introduced to the concept of humanness to avoid any biased answers.2
The questionnaire’s first section was based on Boom and Pennink’s questionnaire (2012) including Sigger et al.’s (2010) measurement tool to calculate the degree in humanness and Scholtens’ (2011) adaptations and connection to knowledge-‐sharing (Lin et al., 2009). A five-‐point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree) with questions being shown in random order was utilized to measure the presence of humanness and the attitude towards knowledge-‐sharing. For both concepts all items per dimension were added up and then divided by the total number of items to determine the mean for each dimension. The total mean scores for the two overall concepts were calculated equivalently. The following rating regarding the degree in humanness and knowledge-‐sharing was used: (1) 2.4 or less: low level; (2) 2.5-‐3.5: moderate level; (3) 3.5 or more: high level (Scholtens; 2011; Sigger et al., 2010).
2 The following paragraphs explain the scales in more detail. Please refer to the Table of Operationalization in Appendix A for an
overview of all the items used in the questionnaire.
21
The questionnaire’s second part used an adaptation of Taylor and Murthy’s (2009) knowledge-‐sharing scale, which enabled the research to add the dimension internal networking behaviour to the existing research model regarding the relation between humanness and knowledge-‐sharing. The study extended Taylor and Murthy’s (2009) research as the present questionnaire related to internal social networks and thus investigated intra-‐organizational knowledge-‐sharing networks compared to electronic networks of practice as in their research (2009). Again, a five-‐point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree) with questions being shown in random order was used to measure the attitude towards or the frequency of a certain behaviour.
The questionnaire’s demographic section served to obtain information about the participants’ social value orientation. It was measured by asking the participants about the federal state of their highest education and of their primary work as people’s values are constructed and reconstructed to a great extent through social structures and interactions and especially so in the context of the latter two (e.g. Bourdieu, 1990; Bourdieu & Wacquant, 1992; Chow et al., 2000; Friehe & Mechtel, 2014; van Lange et al., 1997). To assign a social value orientation category to the participants, they were divided into East and West Germany according to the following criteria: (1) East Germany: education East and primary work East; or education West and primary work East; (2) West Germany: education West and primary work West; or education East and primary work West.
3.3 Validity and Reliability
Both measurement scales used in this study had been utilized in previous studies and thus been verified and confirmed by several authors (Boom & Pennink, 2012; Scholtens, 2011; Sigger et al., 2010; Taylor & Murthy, 2009). The conduct of a factor analysis was therefore not necessary. Nonetheless, Cronbach’s alpha coefficients, means, standard deviations and normal distributions were calculated to confirm the reliability of the measurement tools within the new context.
TABLE 2:
Cronbach’s alpha coefficients for the humanness and knowledge-‐sharing dimensions
Second, as seen in Table 3, the means and standard deviations for each dimension and the overall two concepts were calculated. The means were generally rather high, indicating a high level in humanness and knowledge-‐sharing within the German field of accounting. The standard deviations ranged from .443 to .878, demonstrating that all answers were closely distributed around the mean.
TABLE 3:
Means and standard deviations for humanness and knowledge sharing dimensions
Third, it was checked if the data was normally distributed. One participant might have acted as an outlier; nonetheless there was not sufficient evidence to exclude the participant from the data analysis. Based on the p-‐p and q-‐q plots, normal distribution prevailed for each dimension and the overall two concepts. The moderator social value orientation however lacked a normal distribution. The sample was limited to availability and
23
mirrored the German population rather than being equally divided into an East and West German orientation (Statistical Offices of the Länder and the Federal Statistical Office, 2014): the sample consisted of 9 East Germans and 90 West Germans. It was acknowledged that the model thus suffered in reliability; nonetheless the study proceeded as possible soft indications were still expected.
As the research added the knowledge-‐sharing dimension internal networking behaviour including 17 items to the existing research model, this dimension’s reliability and validity had to be tested. Principal component factor analysis was used to determine whether it was possible to reduce the various items into a littler cluster of components. Varimax rotation was applied as the different factors were assumed to be uncorrelated. Taylor and Murthy (2009) had already identified reputation, altruism and commitment as distinct factors. Through two principal component factor analyses the present research found participation (KMO=.610; p=.000; CV1=.708; CV2=.757; CV3=.441) and helpfulness (KMO=.592; p=.000; CV1=.703; CV2=.878; CV3=.633) as two additional factors. Looking at the rotated component matrix of all knowledge-‐sharing items, it was clearly visible that all but three newly added items loaded with an Eigenvalue greater than .400 on a factor standing in addition to the three existing knowledge-‐sharing factors. As this suggested the prove for the new dimension, another principal component factor analysis was performed and showed that the five factors reputation, altruism, commitment, participation and helpfulness all loaded on one factor (KMO=.806; p=.000; CV1=.370 (assumed to amount to the acceptable threshold of .400); CV2=.691; CV3=.591; CV4=.795; CV5=.746), which was then labelled internal networking behaviour. The Cronbach’s alpha coefficient of .853 furthermore suggested a good internal consistency. Furthermore, none of the knowledge-‐ sharing items showed a correlation above .900, implying that all items had measured distinct aspects. It was thus concluded that the new dimension Internal networking behaviour did add value to the concept of knowledge-‐sharing. Nonetheless, the research’s theoretical contribution warrants cautious interpretation, as only 31 of the 100 participants were included in the internal networking behaviour dimension.
3.4 Data Analysis
Before the data was analysed, it was (1) cleaned, (2) transformed, and (3) recoded in Excel (Sue & Ritter, 2007). The crosstabulation in Table 4 summarises the data and shows how many times each possible category combination appeared in the sample.
TABLE 4: Crosstabulation –
Humanness * Social value orientation and Knowledge-‐sharing * Social value orientation
The statistical analysis was divided into three parts. First, the mean scores for the overall concepts and the individual dimensions were calculated to determine the presence of humanness and knowledge-‐sharing. Afterwards, several one-‐way between groups
East%Germany West%Germany Total
25
ANOVAs for humanness and knowledge-‐sharing and the concepts’ individual dimensions were performed to reveal any significant differences between the two social value orientation groups. A post-‐hoc test was not performed, as there were fewer than three groups within the social value orientation.
Second, the relation between humanness and knowledge-‐sharing was analysed at levels 1, 2 and 3 by various Pearson product-‐moment correlations extended by several regression analyses. Figures 4, 5 and 6 complement Figure 1 by displaying the analysis levels 2 and 3. Pairwise deletion of missing data was applied to use as much data as possible and to counteract the fact that only 31 participants were included in the internal networking behaviour dimension. The Pearson product-‐moment correlation coefficient was calculated at level 1, 2 and 3 for the overall two concepts and for the individual dimensions to determine if there was a significant association between the two concepts and to identify its strength. Linear regression analyses were then performed to examine whether the significant relations of the Pearson correlations were causal as to explain any influences and thus to answer hypothesis 1. At levels 1 and 2, various one-‐to-‐one regression analyses were performed between humanness and knowledge-‐sharing as well as humanness and the individual knowledge-‐sharing dimensions. At levels 2 and 3, several multiple regression analyses were performed between the individual humanness dimensions and knowledge-‐ sharing as well as its individual dimensions.
FIGURE 4:
Level 2 – Relations between humanness and the knowledge-‐sharing dimensions
