Network Analysis in the Information Systems Management: Implications for a Transdisciplinary Approach

Abstract

The Network Analysis in organizations made in last years some meaningful results owing progress in technology and in the approach to organizational networks. The chapter enhances the need to transfer some results of network analysis from management to the analysis of technical networks. Coming from results in the studies, theoretical and empirical, on business ties and on strong and weak ties connected to the mechanism of control, the chapter proposes a transdisciplinary approach to interpret the differences and the evolution of the types of networks through cycles of simplification and complexification of the control systems. Particularly, as results are connected to the adequacy of control tools, it is relevant to consider managerial concepts, such as the span of control defined as the number of subordinates of a hierarchical position, and the connected capability of networks to maintain control, particularly when the system is wide and highly interconnected.

1. Introduction

According to transdisciplinary approach, the introduction of new perspectives needs the definition of the scenario in which hypotheses are located and induce new tools of analysis.

Continuous progress in technology and organizational work approach led to an overlap of the concepts of network and organization. The network seems to be a modern expression of the organization, and the organization increasingly takes on the nature of a network.

The purpose of this chapter is to discuss the results of the network analysis of organizations and the relevance of the organizational concept in the meaningful understanding of networks and their control systems, which is not limited to a pure application of increasingly sophisticated algorithms but seems to become distant from the practice of organizations [1].

The perspective assumed is transdisciplinary, according to the distinction claimed by some authors, regarding its difference from multi and interdisciplinary perspectives [2]; as it implies the use of a model having the organizational matter as basic but with a reciprocal influence on information knowledge about networks.

Network analysis, in the last few years, has undergone a hyperbolic development of digitization and the creation of sophisticated applications [3]. However, this development did not consider the change in organizational conditions in which the situations are analyzed and enormously implemented in the last decades with the complexity of networks [4].

In other words, the hypothesis is that the qualitative side of this evolution has been brought about by the evolution of the network analysis that does not consider the evolution of the networks themselves as products of organizations, their feedback regulations [5], or the myopia of individuals in considering network performances [6]. Consequently, the field of analysis to which we apply is not thought to have changed, while practitioners observe this phenomenon and recognize that it is progressing rapidly [7], in a parallel way with the implementation of transdisciplinary research process [8].

The qualitative aspects of this evolution are striking compared to the quantitative aspects derived from the hyperbolic growth of parameters and indices used to measure the performance of networks. This evolution foreshadows not a simple adaptation of the reference models but a real phase shift [9].

This last step, still in progress [10], represents the limit of the complexification of models built to interpret situations and prepares for the next advancement by exceeding the threshold between two different phases and collapsing pre-existing models in favor of a newer, simpler model [11].

With the organizational evolution, this cycle of simplification that increasingly moves toward complexification and vice-versa, represents the organizational engine of the evolution of the analyzed social and corporate entities, according to the Kuhn hypothesis on scientific revolutions [12, 13].

Far from providing a definitive answer, this chapter represents an invitation to orient research toward the analysis of the evolution of information systems from an organizational viewpoint, considering strictly technical parameters, and broadening the consideration to organizational aspects in a transdisciplinary approach [14].

Different approaches were devised to give a transdisciplinary perspective to the theme of network analysis, already extensively treated in a traditional key. One of these, considered appropriate, is to define the general reference model for which to carry out the analysis [15].

In the specific case, the network analysis can be referred to as a model to which various interacting aspects refer, ranging from the classification of the types of networks to the simplification/complexification cycles of production processes, attention paid to details or to the general framework, and definition of intrinsic/extrinsic quality to end up in the control area that the network structure is able to express (Figure 1).

Figure 1.

Components of transdisciplinary model for network analysis.

2. The first step of the analysis: simplification and complexification cycles

Considering the network to be a universally diffused model structure, we examine the main typologies of macro structures that make more evidently the form of governance (Figure 2).

Figure 2.

Basic structures of network governance/dimensions.

In a basic evolutionary model, network macro structures span from the largest organizations (measured by the number of employed people), structured as autocratic networks, to smaller organizations that can employ as few as a single person managing activities.

Once established, the decentralized network produces an increase in networking among its units and, in the next phase, the emergence of a leader organization, which centralizes some strategic functions dominating the network. This step just precedes the creation of an integrated network, restarting the cycle.

The existence of different network typologies is accepted by organizational theory [16]. The novelty comes from the studies on the motivation for these differences and of their creation.

Starting from these organizational typologies, (Figure 2) a hypothesis on evolutionary model of macrostructures takes as an engine of change the continuous simplification or complexification cycle, which leads to the intrinsic quality of the process/product from one side and an extrinsic quality referred to as the network organization as a whole from the other side [17] (Figure 3).

Figure 3.

Evolutionary cycle of network typologies.

According to the vision of this model, studies on the behavior of systems from the perspective of greater efficiency of performances consider that, when environmental conditions highlight the limits of a greatly integrated and autocratic organization, the drive toward decentralization begins, finding application in the fragmentation of the production process. Its extreme application uses point analysis and a focus on quality because of single actions [18, 19].

This process is achieved by reducing the field of attention of analytical theories and methodologies that constitute a single package of phenomena that leads to a decentralized network as part of the simplification cycle [20].

With the rationalization of production and its greater repeatability, this process eventually facilitates the transition to the outsourcing of product components in external business units that become increasingly complex. External units result in greater autonomy from the parent company, favoring decentralization. Consequently, the evolutionary process moves toward a participatory network organization with specializations related to outsourcing relations centered on the production supply [21].

This evolution led some of these companies to switch to the production of increasingly complex components, allowing some of them to assume pre-eminent dimensions using positions of competitive advantage. In this phase the attention is focused on the extrinsic quality of the production, referring to the company’s overall organization and the local environment from which it draws its culture. This process highlights the cycle of complexification with an expansion of the field of attention, applying synthetic methodologies, such as field analysis, and triggering a process that will again lead to centralized organizations [22].

The idea is to apply this interpretative model to the network, in general, considering that the context of a network is not strictly referred to a business organization but can be extended to a family, a condominium, a neighborhood, a city, a state, or the whole world. Once the network elements have been referred to as coherent, understanding models, the analysis can focus on the links between elements [3].

3. The order of ties

The distinction between direct (or first-order) and indirect ties considers two possible structures of a simple network (Figure 4) related to the question of governance and performance efficacy. In the network in Figure 4b, which contains the same number of elements as Figure 4a, indirect ties are limited to the second order (second-order links). However, in a complex network, the number of indirect ties could increase to a hyperbolic order.

Figure 4.

First- and second-order ties (dashed lines): (a) first-order ties; (b) second-order ties.

Research conducted by the author evidenced that, in networks, the number of first-order ties changes according to the increase of components in a linear way and can be related to their intensity, represented in business organizations by the company turnover (Figure 5) [23].

Figure 5.

Curves of numerousness (x and y) and 6-year frequency of business-to-business ties.

The two curves reveal that most of the intensity of ties, represented in business organizations by the turnover company (y curve), is concentrated on a limited number of business-to-business (B2B) relationships.

In terms of the network, the concrete dynamics of relationships derived from human behavior demonstrate that most of the ties are related to desultory relationships, whereas only a restricted number of links are connected to the highest intensity of relationships [24]. These trends, empirically detected in a sample of Italian small and medium enterprises (SMEs), were confirmed by other researches on supply chains and network management [25].

Particularly, the trend of B2B is recognized as strictly related to the customer and supplier selection process. This novelty, together the inadequacies of the current studies on the topic, orient the future direction of research on supply networks [26, 27].

Similar trends were detected on links connected to social networks in which Granovetter introduced the concept of strong and weak ties with particular attention to economic and organizational links and, in a wider sense, to the social system [28]. Weak ties in social systems are defined as most influential links not reinforced by mutual friendships. In contrast, strong ties are supported by direct, emotional links derived from the custom of relationships.

In the curves, in Figure 5, regarding the frequency of B2B ties, the distinction between the two categories is illustrated. Furthermore, the extension of the examined strength of the links to the structures of the technical networks is reasonable when they are applied to a great range of individuals, organizations, and urban structures [29].

Accordingly, in the research on their behavior, the analysis of link strength is applied with particular attention regarding the analysis of the link intensity according to the distribution of first-order ties [30].

4. The second-order ties

Second-order ties imply the existence of a hierarchical position from which the link is connected in an exclusive derivation or sharing with other elements directly connected to junction or branch points. This structural condition, displayed in Figure 3b, leads to the principle of “Graicunas Span of Control” (GSOC) [31] in the organization of the network, which Graicunas shared with Lyndall Urwick, one of his best estimators and master of scientific management in 30 years [32]. This principle was also listed in the “Ten Principles of Management” recently reconsidered [33]. The GSOC principle defines seven as the number of direct subordinates a manager can adequately manage.

In organizational structures, this problem led to the assumption of a hierarchical-functional order and authority managed with intermediate positions whose number was related to the organizational dimensions and whose order depended on the dimensions of the organizational network and on the managerial structure [34].

This leads to the consideration that, while the number of network connections, which a position oversees, can theoretically be infinite, in practice the organizational performance control capabilities are limited.

Figure 6 shows the hyperbolic growth of second-order ties correspondent to the first ones as calculated (Table 1) by Graicuna’s Equation [Eq. (1)] in which R is the number of second-order ties and n is the correspondent number of first-order ties.

Figure 6.

Trend of first- and second-order ties according to the Graicunas Eq. (1).

First order Ties	Second order Ties
1	1
2	6
3	18
4	44
5	100
6	222

Table 1.

Trend of 2 ns order ties according to Graicuna’s Eq. (1).

$$R=n\left[\frac{2n}{2}+\left(n-1\right)\right]$$E1

After its codification in organizational principles, GSOC was submitted to several critics and cited by Simon [35] as an example of an inconsistent criterion, at best considered a proverb. Particularly, Simon underlined that the GSOC has a relative appliance, as it is adequate only to restricted situations in which other elements, such as technology, information systems, and the managerial process, allow its appliance [36].

Considering the connection between March and Simon, we note that in the Handbook of Organizations edited by March [37], the principle was mentioned and significantly quoted by Urwick [38, 39] for any case limited to closed systems [40].

The last topic in the evolution of networks brings the Graicunas principle back up to date with the limits of control systems in the face of the growing complexity of organizations and with an extension to processes of its interpretative potential.

5. GSOC and control processes

GSOC was examined concerning the measuring of hierarchical dimensions [41] in connection with the order of the hierarchy. Other authors observed that the GSOC had a relative relevance in connection with the simplicity of jobs, increasing its measurement while decreasing with the surge of complexity [42] and speculated on its connection with the spans of accountability, influence, and support. These paradoxes can also occur in simpler scenarios where the invasive network control process coexists with the continuous looming risk of collapse of the control systems [43].

The extension of the GSOC principle to procedures can be proposed by applying a different approach in which the communication process between different organizational positions is intended as a process comprising many steps (Figure 7). The number of these steps is related to the efficacy and efficiency of the control [44].

Figure 7.

(a) Basic process of control compared to the (b) multistep process typical of complex networks.

In the networks, the existence of many points or steps of control may be interactive, involving many steps and operations (Figure 7b) [45]. Consequently, the information processed in the network is subjected to the loss of components and the acquisition of others in continuous renewal (Figure 8), as in computational transfer [46]. This process includes the risk of uncorrected information and inaccurate control feedback.

Figure 8.

Process of information renewal.

Progress in supply chain management implemented innovations in the tracking system of goods and services, producing advancements in integrating operators and software. This progress prompts the question: How many steps could a manager control? How many steps can be detected effectively by an organizational position [47]?

Once more, the attention of scholars and practitioners was focused on the study of hyperlinks [48], in which the multiphase process links two elements from a theoretical viewpoint, while a situation arises in practice in which the two extremes are isolated from each other when steps exceed some number [49].

This reasoning leads to the question, concerning the organizational boundaries of identifiable networks, considering the extension of the relationships between the component elements, and their strength and mutual influence [1]. Consistent with this, the extension of a network that comprises interactive elements can be conventionally established based on the boundary between weak connections subject to maximum renewal and strong connections featuring low renewal. This boundary can be identified while considering the bond intensity at the point of intersection of the x and y curves in Figure 5.

This led to the evolutionary model of organizational networks in which the unrestrainable renewal of ties produces a complexification of controls of each specific structure and consequently a structural evolution of organizations [50]. When limits of adaptation were overcome the evolution proceeds by a jump to another kind of structure in a cycle in which the contingent typology of the network is only a step.

6. Conclusions

With these considerations on network analysis, from a transdisciplinary perspective, the meaning of many measurements and indexing aimed at improving organizational performances invites us to consider the limits of the analysis and the complexity of the tools used for this purpose. A limit of the extension of the number of indices focusing on a few shared indices is conceivable in the future of network analysis [51]. There is also a need to assess the simplification of the analysis to originate systems that can be considered, not just on paper, susceptible to concrete and feasible control. Finally, one should consider the limitation of the boundaries of the network, which do not ensure adequate reliability of their consistency if certain thresholds of the degree of interdependence are exceeded, especially concerning control systems.

Conflicts of interest

The authors declare no conflict of interest.

Funding

This research received no external funding.