Roger Clarke's Web-Site
© Xamax Consultancy Pty Ltd, 1995-2018
|Identity Matters||Other Topics||Waltzing Matilda||What's New|
Working Paper for first presentation as an ACIS'15 Keynote
Version of 30 September 2015
Roger Clarke **
© Xamax Consultancy Pty Ltd, 2015
Available under an AEShareNet licence or a Creative Commons licence.
This document is at http://www.rogerclarke.com/SOS/EPinIS.html
When we devise a method to address a particular research question, we think about many things, including the unit of study. But do we think about the perspective from which we observe whatever it is that's within our field of view?
In the majority of mainstream IS research, one perspective dominates. Organisations use IT as a means of intervening into a context. And the behaviour of the phenomena is mostly observed from the viewpoint of those organisations.
Is that viewpoint the only legitimate one for IS research to adopt? Even if it is, are the clients' interests best served by IS research that adopts that perspective alone?
This presentation identifies alternative perspectives. It draws attention to some serious negative consequences of wearing blinkers to restrict our field of view, and proposes that these alternatives need to be defined as being within the IS discipline rather than outside it.
IS academics spend a great deal of their time on the conception, design, conduct and reporting of research. The many aspects that we invest energy into include existing bodies of theory that can be used as the theoretical lens through which to view phenomena, the formulation of research questions, the selection and application of research techniques to answer those questions, and, particularly in scientistic research, population definition, selection of sampling frame and sample, and external validity.
This paper concerns itself with one particular notion that is almost invisible in the research process, almost always left implicit. It is contended that leaving this aspect festering in a dark corner is a serious weakness in our discipline, that this has led to a number of highly unfortunate outcomes, and that we urgently need to throw light on the notion and address the issues.
Discussions about research frequently refer to the 'unit of study' or 'unit of analysis'. There are many levels of abstraction at which we can observe phenomena around us. We choose to bring our focus to bear on one, which may be, for example, a category of employees or customers, a particular organisational entity, a category of organisational entities, or an industry segment. We then frame our research questions in terms of that particular unit of study, and develop our research design with it in mind.
The invisible aspect of the research process that is of relevance here is related to the notion 'unit of study', but different from it. The term 'perspective' is adopted in this paper.
The term 'perspective', as it is used in this paper, refers to the viewpoint from which phenomena are observed. But the viewpoint is not detached from reality, as is implied by phrases frequently encountered in IS research literature - a scientific, an interpretivist or a critical-theory perspective, a disciplinary perspective, a theoretical perspective. The perspectives of interest are those depicted by the tale about blind men describing an elephant differently, depending on which part of it they are holding. The viewpoints this paper has in focus are practical, empirical or real-world in nature, as seen from the standpoints of stakeholders in the process that is under observation.
Recognition that phenomena look different depending on the vantage-point from which they are viewed has been stronger since interpretivism gained a foothold in IS research. For example "... the intent of [interpretive] research [is] to increase understanding of the phenomenon within cultural and contextual situations; where the phenomenon of interest [is] examined in its natural setting and from the perspective of the participants; and where researchers [do] not impose their outsiders' a priori understanding on the situation" (Orlikowski & Baroudi 1991 p.5, emphasis added). The singular 'perspective of the participants' is presumably an editing error, and should be the plural 'perspectives of the participants'.
There are a number of differentiators of interpretivism from scientism (Walsham 1995, Clarke 2000, 2001). Those relevant to the present discussion are the express recognition that:
For simplicity of expression, it will be assumed during the early part of the discussion that relevant viewpoints are those of particular entities, or categories of entity. That assumption will later be relaxed.
In order to determine what perspective(s) any particular research project adopts, several candidate citeria need to be considered. One possibility is to identify the client(s) for whom the research is conducted. The term 'client' is usually associated with consultancy rather than with academic research. On the other hand, relationships of the nature of contract and patronage are far from uncommon in IS research. An explicit client may exist in the case of commissioned research, or a client relationship may arise because of the source of funding for the Centre or research program under whose aegis the work is performed. Such arrangements may not be directly determinative of the design and conduct of research, but they influence the manner in which IS researchers conceive, execute, and write up their work. Many funding sources, on the other hand, such as government research funding agencies and philanthropic organisations, may be primarily concerned with factors such as quality, and may be neutral in relation to the perspective adopted.
A second candidate criterion is the audience(s) to which reports about the research are addressed. The convention in the IS research community is to identify both 'implications for research' and 'implications for practice'. The key question is the category/ies of reader that the researcher has in mind as they frame their comments relevant to practice.
The third candidate criterion is the entity/ies or category/ies of entity that the researcher envisages as the beneficiary/ies of the work. Some research is 'pure' or 'fundamental'. Some is concerned with research methods, and some with aspects of the IS discipline, such as performance measurement and controls over malpractice such as plagiarism. The majority of IS research, however, is targeted at one or more real-world beneficiaries. The beneficiary criterion appears to provide the most appropriate approach for determining the perspective adopted.
It is useful to revisit the concept of 'unit of study', and to distinguish it from 'perspective'. Unit of study refers to the object of the verb 'observe'. Hence 'a researcher observes phenomena at the level of abstraction called unit of study'. Perspective, on the other hand, is part of the subject of the sentence, as in 'a researcher adopts a perspective in order to observe phenomena at the level of abstraction called unit of study'. (This approach can be further expanded in order to show the relationship with other aspects of research, e.e. by appending 'using a particular theoretical lens to impose order on the expression of observations').
Searches for the term 'perspective' in leading IS journals identifies limited usage in the sense described here. It is frequently used to refer to the theoretical lens, but seldom to the interests of a stakeholder. Searches for other terms associated with this concept were also largely unsuccessful, in both text-books on research and in journal articles dealing with research process. Remarkably, the notion appears to be largely absent from the debates about the 'core' of the IS discipline stimulated by Benbasat & Zmud (2003) and Walsham (2012).
Perspective may be made explicit, but, even if implicit, it has a very significant influence on the entire research undertaking. As we frame our research, we are adopting a particular point of view, our formulation of research questions is strongly influenced by that view, we evaluate alternative designs using criteria shaped by that perspective, and what is included within and what is excluded from the potential outcomes of the research are oriented towards the interests of the chosen beneficiary, and expressed in ways intended to communicate to the chosen audience.
The following sections articulate the basic concept, and draw out its implications. The next section identifies the particular perspective that dominates IS research. Alternative perspectives are then considered, and their legitimacy evaluated. Approaches to research that actively reflect more than a single perspective are discussed. Implications for the discipline are identified.
Casual observation of the content of IS journals and conference proceedings over a long period has given the author the strong impression that the mainstream perspective adopted in IS research is that of the sponsor of the information system. By 'system-sponsor' is meant the particular organisation that is developing or implementing the system, or causing it to be developed or implemented, or for whose benefit it has been installed. Surveys and experiments are conducted in order to understand the impacts of interventions. A great deal of of interpretivist research is performed in order to provide broader understanding of those impacts. In design science, the notion of design is "the purposeful organization of resources to accomplish a goal" (Hevner et al. 2004). It literally 'goes without saying' that the goal is formulated in order to serve the interests of the organisation by whom or for whom the intervention is made, or the design activities are performed.
In some cases, the perspective adopted is that of a category of organisations, such as corporations installing an ERP package or adopting an ERP cloud-service, or hospitals implementing an electronic health record system. In joint ventures and collaborative inter-organisational schemes, the system-sponsor may be a collective. In some cases, the system-sponsor may be the client. In the large majority of cases, the system-sponsor is the, or at least the primary, beneficiary, and the, or at least the primary, audience to which implications for practice are addressed.
Organisations generally, but especially business enterprises, define objectives solely in terms of the organisation's interests. This is 'in the DNA', because the nature of the joint stock company is such that directors have a legal obligation to act in the interests of the company. This mode of thinking has migrated beyond the business context, because mission statements and corporate objectives are now the driving force in public sector and non-profit organisations as well.
The system-sponsor may recognise that other entities also have interests. But those interests are seldom treated as objectives, and operate instead as constraints. And, even then, only those entities are considered that are perceived to be sufficiently powerful that they need to be recognised as stakeholders. Legal requirements and softer forms of influence such as industry codes and Ethics Committees also represent constraints on organisations' freedom of action. However, the last few decades have seen significant winding-back of 'regulation' in favour of mere 'governance', and hence these constraints are in many circumstances very weak. 'Self-regulation' meanwhile is not just the softest form, but is also largely an empty promise by all but the few most strongly values-driven organisations. The 'business ethics' and 'corporate social responsibility' notions are excusatory, designed to keep the threat of regulation and enforcement at bay, not to permit the interests of other parties to become part of the organisation's mission and objectives.
The adoption of the system-sponsor's perspective is clearly both legitimate and valuable. Moreover, it is entirely appropriate in the case of Management Information Systems (MIS) in the US tradition, and Wirtschaftsinformatik (WI - Business Informatics) in German-speaking countries. On the other hand, the broader fields of Information Systems (IS) and Information Management (IM) could reasonably be expected to be more diverse, and to encompass research that adopts perspectives other than that of the system-sponsor. It appears, however, that IS and IM have progressively also adopted the norms that are mainstream in both business schools and software engineering schools.
A later section moves beyond the casual observation on which the analysis in this section was based, and considers some empirical evidence. First, however, it is necessary to identify perspectives that could be alternative, competitive or complementary to those of the system-sponsor.
The IS discipline emerged in the mid-1960s, particularly in Scandinavia, the UK, the US and Australia (Clarke 2008, Hirschheim & Klein 2012). This followed the application of computing to administrative tasks, which had commenced in the early 1950s in the UK and the USA. The discipline began by adopting and adapting prior work of relevance, from such areas as Organisation & Methods (O&M), and pre-computing Automated Data Processing (ADP). The initial context was computer-based systems, and the initial challenge was to find ways to manufacture software that enabled systems to perform administrative work (very) quickly, but also accurately. So it was natural that there be a strong focus on the interests of the system-sponsor.
However, during the following quarter-century, the impact and implications of IS expanded enormously, as computing progressively married communications to produce IT/ICT, and as the scope of applications extended to all organisational functions, and then to inter-organisational systems. By 1990, extra-organisational systems had emerged, in such form as videotext, bulletin-board systems, ATMs and EFTPOS, which actively involve not only large organisations but also individuals and micro-businesses (Clarke 1992).
A variety of movements have endeavoured to achieve broader horizons for IS research and practice. The scope of 'system' is far broader than computer-based processes alone, and both human-computer interface and human-performed processes have been incorporated within the discipline and profession. Some approaches have emphasised the interests of individual users within the organisation, and proposed a participative approach to the analysis of requirements and the design of systems (e.g. Land & Hirschheim 1983). As systems reached beyond organisational boundaries, organisations other than the system-sponsor were recognised as having a stake in the design and implementation. As individuals outside the organisation became users of networked computing facilities, their actions and their interests also needed to be encompassed within the frame of reference. As computing became a major consumer of electricity, and as cathode-ray tubes mounded up in scrapyards, impacts on the environment came to notice, and needed to be defined as being within the scope of the IS discipline.
In the contemporary context, a range of possible perspectives can be identified. Any discipline must conduct some research that is internally-facing, such as discussions of scope, of method, and of quality assurance mechanisms. The IS discipline may be seen as having published an excessive volume of papers whose beneficiary and audience is itself. This has been partly because of its eternal existential angst, but also because, in many segments, rigour has been permitted to overtake relevance. Such genres as the design and validation of survey instruments, and reports on minute variations to prior research, are neither accessible to professionals nor of any use to them. The following sub-sections focus on literature in which relevance to IS practice is the primary criterion and rigour is a vital characteristic rather than the primary objective.
A variety of approaches have been adopted to distinguishing categories of subject-matter. The approach adopted here has some similarities to that in Keller & Coulthard (2013), but differs in a couple of important respects. Three dimensions are identified, each with different levels of abstraction within them. The first - which has already been introduced as including the dominant perspective in IS research - is referred to here as the economic dimension. The two further dimensions are the social and the environmental.
Human activity is coordinated and directed by nested layers of institutional arrangements. Phenomena can be observed from the viewpoint of any of those levels of abstraction. In order to distinguish these viewpoints from the alternatives that follow, the dimension is referred to here as economic.
The previous section identified the interests of individual organisations, or of categories of them, as being the dominant frame of reference evident in IS research. It is also feasible to undertake research from the viewpoint of sub-structures within an organisation, such as work-groups, branch offices and subsidiaries. Above the level of the individual organisation, there are strategic partners, industry segments, value-chains, regional economies, nation-states, supra-national regions (such as the EU, NAFTA and ASEAN), and the world economy. Although they represent a minority, a range of IS research projects adopt such perspectives, particularly those of clusters of strategic partners, and industry value-chains.
Each individual, as a person rather than as a labour resource within economically-oriented organisations, has interests that are affected by information systems. In principle, each person has a perspective from which phenomena can be observed, although in practice it is necessary to consider categories or segments of people.
The idea of 'person' needs to be unpacked, however. It is conventional to apply the term 'user'. Research needs to be conducted from the perspectives of the users of eCommerce and MCommerce services, government services, Internet Banking, games, SaaS, social media, media and entertainment, etc. Segments can also be recognised that have specific needs, such as the physically-handicapped and the socio-economically disadvantaged. Examples of research questions at this level of abstraction include:
What are the characteristics of messaging, workflow and calendar services that ensure a fair work-life balance for employees?
What contract terms do consumers need from cloud services such as social media and family-tree web-sites?
In addition to users, there are also many people who are not directly involved in information systems, yet are affected by them. Such role-categories as job-applicants, credit-applicants, tenants and surveillance subjects are usefully referred to as 'usees' in respect of background databases that are consulted by commercial and administrative decision-makers (Clarke 1992). Research needs to be conducted from the perspective of usees, with other actors, including the system-sponsor, identified as stakeholders, and their interests treated as constraints, not objectives. An example of a research question at this level of abstraction is:
What threats to tenants' interests arise from commercial databases that gather and publish data about tenants, and what safeguards, mitigation measures and countermeasures are available to tenants to counter those threats?
At a higher level of abstraction, a range of institutions exists, within which human activities are coordinated for social rather than economic purposes. People cluster into family, kinship and household groups. They also form geographical communities such as villages, precincts and suburbs, and common-interest communities that may have a physical locus and/or take the form of virtual organisations. They may or may not be incorporated, e.g. as a corporation, association, cooperative or club. They may arise from a commonality of geography, ethnicity, language, religion or perceived interests. Their purpose may be of an entertainment nature, or they may have a productive, service, representative or advocacy function. An example of a research question at this level of abstraction is:
What are the key requirements of a peer-to-peer scheme to accumulate and publish information about the service-levels provided by a particular government agency?
At an even higher level of abstraction, we conceive of societies and populations, in some cases within geo-political boundaries, in other cases crossing national borders. Most broadly, we can conceive of the perspective of 'world society', or humanity as a whole.
Is it appropriate that the kinds of research questions that arise along the social dimension be investigated within the IS research community?
A coinventional normative standpoint is that interventions of all kinds are intended to be for the benefit of the human race. That assumption is being questioned, with propositions that other forms of life and the planet as a whole need to be included within the objective function. The IS discipline must at least recognise the existence and importance of that debate, and arguably should participate in it.
Ecologies need to be recognised, again at various levels of abstraction. At a low level, the impact of CRTs in landfill, and of electricity consumption by computing equipment, can only be discussed if concepts such as individual-species-within-context, local bio-communities, regional bio-communities, or the biosphere as a whole, are regarded as a set of interests from whose perspective phenomena can be observed.
All of the examples given in the two preceding sub-sections involve reasonably definable entities, whether associated with a physical person or with some fiction such as a joint stock company that is widely-recognised, and in some cases even assigned legal rights and responsibilities, such that the lack of a physical existence is overlooked by common consent.
Reification also arises in the environmental dimension. For example, 'primitive' societies are recognised as having close associations with nature, as in the specifically Australian Aboriginal notion of 'country'. Echoes reverberate within 'advanced' societies, such as the not-quite-yet-archaic notion of 'Mother Nature', and the modern notions that are complementing and replacing them, such as nature and wilderness reserves, national parks and environmental trusts. However it is far less accepted to imbue such ideas with the trappings of an entity, as is done with corporations, nation-states and incorporated associations. An example of a research question at this level of abstraction is:
What are the key factors to be considered when evaluating the sustainability profiles of alternative designs for information systems?
Beyond ecologies in water and at ground-level, climate change issues cannot be meaningfully discussed unless that part of the earth's atmosphere is considered in which weather happens - the troposphere - and it is treated as though it were an entity with a point of view. More abstract notions on the same dimension include the planet - or, more poetically, Spaceship Earth. A research question at this level of abstraction is:
How can the eCommerce theory be most effectively applied to carbon trading?
Again, research from these perspectives is essential. If IS is not the appropriate discipline to consider the environmental impacts and implications of information technologies and systems, which discipline(s) do we defer to?
In Table 1, some relevant points are identified on each of the three dimensions discussed in the previous section. In order to gain some insight into the extent to which these alternative perspectives are legitimately adopted in IS research, this section adopts two approaches. The first sub-section considers a particular form of research question. The second sub-section reports on an empirical assessment of IS literature in order to assess the extent to which the various perspectives are and are not already represented in venues for the reporting of IS research.
|World Economy||Humanity||The Planet|
|Supra-National Region |
(e.g. EU, NAFTA)
|Nation-State||A Society||The Troposphere|
|Sector / Value-Chain||A Community||The Biosphere|
|Organisation||A Person||A Localised Ecology|
In order to provide a concrete test, the following research question was contrived:
What are the impacts ... of the withdrawal of the customer option of receiving printed invoices through the post?
The base case involves replacing the ellipsis "..." with "on an organisation". This adopts the conventional perspective of a corporation or small business enterprise, and appears to be a question of a kind that is mainstream in IS research.
Applying even the points on the three dimensions listed in Table 1 above, over a dozen further variants of the research question can be readily generated, each of whose legitimacy as an IS research question needs to be considered. For example:
What are the impacts, ON THE REGIONAL ECONOMY TO WHICH THE ORGANISATION CURRENTLY OUTSOURCES INVOICE PRINTING AND MAILING, of the withdrawal of the customer option of receiving printed invoices through the post?
What are the impacts ON PEOPLE WHO HAVE NO INTERNET CONNECTION of the withdrawal of the customer option of receiving printed invoices through the post?
What are the impacts ON FORESTS of the withdrawal of the customer option of receiving printed invoices through the post?
It is entirely tenable for these questions to be regarded as intellectually uninteresting, or insufficiently rewarding, or not readily researchable using the techniques commonly applied by the researcher or the institution they work within, or simply lower-priority than other competing topics. The question here, however, is whether such research questions in principle fall inside or outside the scope of the IS discipline.
The author's working definition of the scope of the IS discipline has long been "the multi-disciplinary study of the collection, processing and storage of data; of the use of information by individuals and groups, especially within an organisational context; and of the impact, implications and management of artefacts and technologies applied to those activities" (Clarke 1990). The author contends that all of the research questions in this test case fall within the scope of that definition, and hence are appropriate subjects for IS research. Other members of the discipline may, however, disagree, or may feel uncomfortable having such broad responsibilities thrust upon them.
Reference was made above to the author's strong impression that the mainstream perspective adopted in IS research is that of the sponsor of the information system. In order to test that inference from casual observation, several venues were inspected and the perspectives adopted by authors sought out. It is unusual for perspective to be formally declared by authors, and in most cases it was necessary for it to be inferred. A convenient way to present the outcomes of the inspections is by identifying a short-form of the research question that the author addressed.
Within resource-constraints, a modest volume of data was sought, from a spread of sources. The following venues were selected from the three regions adopted by the Association for Information Systems, in each case one journal and one set of conference proceeedings:
In all cases, the Volumes and events of the most recent complete year were selected, which was at the time 2014. This was done despite the longer lead-time from research conception to publication for journal articles compared with conference papers. For four of the six venues, all papers were inspected. In the two largest venues, AMCIS and ECIS, respectively 10% and 5% samples were pseudo-randomly extracted, in order to provide article-counts similar to those in the other venues.
For the purpose of this targeted analysis, each paper was classified by the author into one of seven categories. Table 2 shows the categories, and the aggregate counts. Appendices 1-6 list the research questions in each of the categories of greatest relevance. In many cases, the research question was explicit, and in most others it was reasonably clear. In a range of circumstances, it was necessary to read into the paper in order to ensure appropriate categorisation, in particular to identify who the researcher perceived as being the beneficiary of the work. Some editing of questions has been performed, to ensure readability when lifted out of context, and to shorten unduly long research questions.
As Table 3 shows, the aggregate figures in Table 2 mask material differences in the distributions across the categories in the various venues. Nor is there any simple pattern based on either the region or the kind of publishing venue. Generalisations from the data accordingly need to be particularly cautious.
In order to draw conclusions about the degree of system-sponsor dominance, several measures could be used. The most liberal interpretation is that in only 79 / 119 papers which observed human phenomena was the system-sponsor's point-of-view unequivocally adopted (66%). At the other extremity, it could be interpreted as being all categories except discipline-internal topics, social and environmental (91%). This does, however, include 12 %age-points where some other-party perspective was apparent. Two examples in this category were 'How can a feedback engine for a self-care app be designed?' and 'Does a digital information service in a remote retail store service customers' needs?'.
It appears to be both reasonably clear and reasonably consistent across all venues that very little research was reported on the social dimension, i.e. only a tiny percentage of the research is primarily from a perspective that is not that of an economically-oriented organisation. Examples included 'How can social networking transform the culture of a developing country?' and 'How to design a bedside device to promote healthy sleep patterns for children?'. No research on the environmental dimension was detected in the sample. The sole contenders were 'What factors influence Green IT assimilation?' and 'Who uses the Community-Based Environmental Monitoring system, how, and how might it be improved?'.
Recognition that the system-sponsor is not the only entity whose interests could frame research design leads to a further question. Can IS research be undertaken, at least in some circumstances, which expressly considers more than one perspective?
Aspects of the multi-perspective notion are of course evident in interpretivist research generally: "The core idea of interpretivism is to work with [the many] subjective meanings already there in the social world; i.e. to acknowledge their existence, to reconstruct them, to understand them, to avoid distorting them, to use them as building blocks in theorizing" (Goldkuhl 2012). The notion of a universal perspective is superficially attractive. This would involve identifying all stakeholders, identifying their interests, and observing the phenomena from all of their perspectives. However, such a quest would be at least herculean, and arguably god-like, because of the enormous range and diversity of parties who may perceive an interest in the relevant phenomena, and the relativity of their perceptions.
There is scope, however, for intermediate approaches. Some research adopts the perspectives of two stakeholders or stakeholder categories. It will commonly be the case that such research will encounter conflicts between the two sets of interests. A descriptive approach would result in 'compare and contrast' of the two views. For example, insight into user interfaces or business processes can be gained by comparing the designer's intent with the users' interpretations and practices. Similarly, an intervention could be evaluated from the perspectives of both the system-sponsor and the users, resulting in a research question such as:
What are the impacts ON THE ORGANISATION, AND ON CUSTOMERS, of the withdrawal of the option of receiving printed invoices through the post?
Alternatively, the impact could be observed not from the perspective of some generalisation of 'customers', but rather from the viewpoints of various categories of customer, such as those in remote areas, those with limited IT skills, and those in the lowest socio-economic groups. Such a multi-perspective study may be challenging to report in a conference paper. However, none of the 129 journal articles in the sample adopted such an approach either.
A particular context in which considerable effort was invested during the era of early inter-organisational systems was international trade EDI. This was appropriately modelled as a constrained network rather than a supply chain, and was in many countries characterised by the absence of any organisation sufficiently dominant to be capable of dictating the system architecture, the commercial relationships, and the priorities. The preferred approach was described as 'win-win-win', i.e. each entity needed to perceive that their interests were served (Cameron & Clarke 1996). Hence, an example of a research question framed in terms of multiple perspectives would be:
What are the critical success factors for a project management framework for collaborative inter-organisational systems, from the viewpoints of each of the main players?
Within the sample of 253 papers discussed in the previous section, the sole example that opened up such possibilities was 'How can we enhance information flow within the agriculture domain?'.
A recent example of multi-perspective IS research is Agarwal et al. (2012). This examined cyber-collective social movements (CSMs) such as the use of social media in the 'Arab Spring' of 2010-12. After surveying the available research methods literature, the authors developed an analysis based on Individual Perspective, Community Perspective, and Transnational Perspective. The paper featured no system-sponsor, but rather three levels of abstraction of the social rather than of the economic dimension discussed earlier in this paper. This is also noteworthy in being one of only a small number of papers that were found in which the term 'perspective' is used in a manner similar to that proposed in the present paper.
The analysis of perspectives in IS research undertaken above has further implications. This section considers the extent to which the perspectives of various actors involved in IS lead researchers beyond the descriptive, explanatory and predictive, towards the normative. The sub-sections below consider three distinct approaches.
The term 'constructivist' can be applied to such techniques as action research and particularly design science (e.g. Iivari 2007). A constructive approach to research shifts beyond the merely descriptive, explanatory or predictive. Design science embodies the normative mode, at least implicitly and often expressly. For example, applying Guideline 3 of Hevner et al. (2004), a suitable form for a research question using a design science approach is:
What is a feasible, good design [for a particular system] that can be implemented in the business environment?
Yet, whereas Iivari regards action research as having a 'client', he does not acknowledge that design research generally does as well. The practice is such that the perspective and values that design science approaches adopt are usually those of the system-sponsor, with the interests of other actors relegated to constraints, sometimes of a political nature (i.e. recognising those with power as stakeholders), but often merely ethical and therefore lacking in motive force.
We discuss ontology (existence), epistemology (knowledge) and methodology (process), but seldom teleology (the purpose for which we apply those ideas). In one of the few articles found that address such matters, Introna (1996) equated teleology in IS with management-driven approaches that overlook the "social reality of the information systems, people and organizations". That dominance of the system-sponsor's perspective has been sustained throughout the two decades since then, to the extent that it is implicit and seldom brought to the surface.
The preceding discussions of alternative perspectives, and of multi-perspective research, give rise to the question as to whether constructive approaches to IS research can adopt perspectives other than that of the system-sponsor. An example of a research question of this nature was addressed in Clarke (2014):
How can consumer-oriented social media be achieved?
This was decomposed into:
* What are the desirable features of consumer-oriented social media?
* What impediments exist to the emergence of such services?
* What means exist to overcome those impediments?
Discussions about research frequently go no further than distinguishing between 'pure research' - undertaken in order to contribute to abstract, theoretical understanding - and 'applied research', which commences with a tool and uses it to intervene within a context. Consideration may be given a priori to the tool's suitability, but only if there is a defined purpose for which the research is being undertaken, and even then the tool's suitability is commonly overlooked or discussed only at the end of the work. The notion of instrumentalist research, on the other hand, does not begin with a tool, but with a perceived problem. The approach is goal-oriented, and involves a search for a solution to that problem (Clarke 2001).
Within the interpretivist approach, it is only necessary to acknowledge the existence of multiple perspectives. The multi-perspective instrumentalist approach, on the other hand, goes further by treating each of the parties as a beneficiary and/or audience, rather than orienting the research question, design, conduct and outcomes to primarily serve the needs of just one of the parties. Hence, for example:
How can an organisation MANAGE the negative impacts ON ALL PARTIES of the withdrawal of the option of receiving printed invoices through the post?
Such a project would need to select and apply principles and a calculus for the balancing of interests and the resolution of conflicts. For example, a state of Pareto efficiency could be sought, in which no party can have their interests any better served without harming the interests of another party. A relatively simple two-party example in the B2B space is Koh et al. (2004), which considered both the customer and supplier perspectives on outsourcing contracts, with particular reference to the mutuality of psychological obligations within the relationship.
It is arguable that design science is at its most effective where it acknowledges the existence of multiple sets of interests, and devises ways to accommodate them. But to date the field has generally adopted a singular notion of beneficiary and/or audience, and defined that party's interests as the objective, and the interests of other parties as constraints. Despite the accumulated understanding of socio-technical thinking (Emery & Trist 1960, Mumford 2000), a multi-perspective 'participative design science' approach awaits articulation.
A research philosophy that is clearly both instrumentalist and multi-perspective is critical theory research. Although the approach is subject to multiple interpretations, it is directly concerned with conflicts among the interests of the various actors, and with the power-structures that determine the outcomes. It also has a strong tendency towards instrumentalism, with the intention of influencing the phenomena that are under observation (Klein & Myers 1999, McGrath 2005).
The discussion to this point has been primarily within what has been referred to in this paper as the economic dimension. The social dimension has a toe-hold in the IS literature (4% of the sample of 253 articles in selected venues in 2014). The environmental dimension, on the other hand, was not represented in the sample at all. It is therefore necessary to consider the extent to which these dimensions should be defined to be within the IS discipline's scope.
The purpose of public policy research is to develop policies, and articulate programs, that are intended to achieve normative objectives expressed in all of social, economic and environmental terms, and that depend on effective cooperation among disparate organisations, usually across the public, private and often also the voluntary sectors.
Research into public policy matters is not well-served by 'pure research' motivated by inquisitiveness, and intended to contribute to abstract, theoretical understanding. Nor is 'applied research' appropriate, because this begins with a tool and looks for things to which to apply the tool. As the saying has it, 'When you have a hammer in your hand, everything looks like a nail'. This is a wholly ineffective way to approach a public policy matter. The driver in public policy research is a problem which includes but extends beyond the economic dimension, and which demands an instrumentalist mind-set, in order to devise possible solutions, evaluate them, articulate them, and implement them. Public policy research commonly requires the engagement of multiple disciplines.
Because of the breadth of both IT and its applications, many aspects of public policy are heavily dependent on IT understanding, and there are few aspects for which IT understanding is irrelevant. The computing and communications science and engineering disciplines are capable of contributing to public policy. However, they are less well-positioned to do so than is the IS discipline, because IS by its nature draws heavily on a range of social sciences and business disciplines, and hence already embodies some of the multi-disciplinary thinking that is essential in the public policy arena.
In practice, however, the IS discipline has largely excluded public policy from its scope. A primary reason for this has been the dominance of the system-sponsor perspective in IS research, as discussed above. Another factor has been the strong leaning towards the economic dimension, almost to the exclusion of the social and the environmental. A third factor has been the perception that the scientism to which IS has long aspired deals only with descriptive, explanatory and predictive modes, excludes the normative, and hence deprecates instrumentalist research.
Instrumentalism and the handling of normative propositions are not confined to constructivist approaches such as action research and design science. For example, some forms of economic analysis, including game theoretic approaches, consider the implications of measures for an economy. However, such research prefers highly rigorous methods, which necessitate the use of models which are highly simplified in comparison with the realities that they nominally represent. Because they control for very few of the confounding variables, they throw very little light on the real world and generally fail the test of delivering useful research outputs.
The proposition that policy is relevant to IS is challenging to the dominant conception of the discipline. If adopted, it would require the co-option of yet more research techniques, many of them dealing with qualitative data, and of necessity lacking the precision and rigour that science-obsessed members of the discipline crave. Evaluation techniques are necessary, ranging from those familiar within business schools - such as the largely political notion of a 'business case', the more economics-oriented cost/benefit analysis, and the broader concept of risk assessment - to others that have hitherto been held at arm's length - such as technology assessment, privacy impact assessment, and their extended forms of surveillance impact and social impact assessment. Further, beyond fringe techniques such as the Delphi method, new data collection techniques are needed that are designed to cope with the normative as well as the explanatory, and the social as well as the economic. Contenders include instrumental futurism (Clarke 1997) and quasi-empirical scenario analysis (Clarke 2015).
Developments in IT continue apace. The current round of excitement includes 'big data', the Internet of Things and its corollary of increasingly intensive 'datafication', digital business and digital users. Inanimate things are now expected to contain embedded chips. Animals increasingly do. Individuals are coming to be perceived as walking, talking data generators, and chipification is no longer the paranoid delusion that the author was accused of portraying in Clarke (1994). A new phase of data expropriation and data exploitation is in train.
IS researchers claim to be at least observers of mighty revolutions, and at least to some extent contributors to them. Can a focus on the economic dimension, and on organisational perspectives, be enough to define the discipline? Or do we, within the IS discipline, need to perform research from multiple perspectives, including those of people as users and usees, and of groups of people in the forms of communities and societies? Further, do we need to embrace research that adopts more abstract perspectives, reflecting the needs of physical ecologies?
The author's contention is that public policy, including its inherent involvement of normative propositions, its instrumentalist orientation, and its reflection of the perspectives of multiple stakeholders, and its inclusiveness of social and environmental as well as economic considerations, is nonetheless a legitimate element within the IS discipline's scope. We have a moral obligation to consider the full breadth of implications of the technologies that we assist in developing and applying and that we subject to study (Clarke 1988). We also need to address the pragmatic or political issue of 'well, if not IS, then which discipline claims the field?'.
This section considers the implications that the analysis presented above has, firstly for the IS discipline, and secondly for IS practitioners.
This sub-section considers three implications, firstly for the quality of research, secondly for individual researchers, and thirdly for the discipline as a whole and its gatekeepers.
It was noted above that the adoption of the system-sponsor's perspective to the virtual exclusion of other perspectives brings with it a number of important corollaries. The system-sponsor's interests represent objectives. On the other hand, the interests of all other entities that are involved represent, at best, constraints. Other entities that have institutional or market power are regarded as stakeholders, such that their interests are factored in, whereas the interests of other entities are largely ignored.
The excluded perspectives are most commonly those of users, usees, communities, and the various levels of abstraction along the environment dimension. And because they are excluded from the design process, they surface during the implementation phase as 'impediments' and 'barriers to adoption'. That results in, at best, the costly retro-fitting of features and 'public education' campaigns, and, at worst, negative return on investment and outright project failures.
The first implication of the analysis of perspectives in IS research is that the entire field of 'impediments' and 'barriers to adoption' is an aberration. Put another way, if the IS discipline's scope and body of knowledge had been well-balanced, there would be no such field of study. By expanding the discipline's scope to encompass user, usee, community and environmental perspectives, all relevant parties will be recognised as stakeholders from the outset, and the necessary features will be part of the design, not expensive afterthoughts. There will be of course be mis-analysis, and there will be ineffective designs, and there will be changes in contexts and expectations; but there will be no more late and unpleasant surprises based on failure to identify and understand the relevant perspectives during the foundation phases of the project.
The author contends that:
That translates into the following obligations on researchers, in order to address the deficiencies discussed in this paper:
Declarations of conflicts of interest are now common, but only in relation to funding sources. On the other hand, the AIS Research Code of Conduct at cl.8, on p.8, does not limit the scope to funding sources, but instead says "Declare any material conflict of interest or relationship that might interfere ..." (emphasis added). Objectivity, to the extent that it may be capable of being operationalised, is most usefully treated as being 'explication of bias'. A client relationship is only one of the possible sources of bias. Others include:
Of course, there are many circumstances in which some aspects of this are already evident to the reader, e.g. from the author's affiliation, from an explanation of the motivation underlying the research, from the choice of theoretical lens, or from the choice of research method. The nature of the declaration may vary, but the author has the obligation to make clear not only the client, but also the envisaged beneficiary/ies and audience(s) of the work.
There are substantial institutional barriers to the broadening of perspectives in IS research, and to the further step of recognising public policy issues as being a legitimate focus. The gatekeepers include editors and program chairs, reviewers, staff selection committees, promotions and review committees, and research funding committees and their assessors.
There are at least a few signs of progress in achieving acceptance. Firstly, there has been very substantial change in conventions over the discipline's 50 years to date. For example:
Business journals such as HBR and the Sloan Mngt Review embrace the normative mode and instrumentalist motivations from a business perspective, and have standing within the IS discipline despite that. In my own recent experience, proposing the topic of 'The Missing Perspectives in IS Research' did not result in a Keynote invitation being rescinded. A Panel Session on 'IS Scholars and Policy' is scheduled for ICIS 2015. And searches in the IS literature for exemplars of policy research do not come up entirely empty-handed.
Notably, Agarwal et al. (2012) differs from mainstream IS research not only in the ways identified above, but also in that the factors under study were social and political, not economic. They fell specifically within the field of policy studies, not business studies. Yet the Editors of a mainstream IS journal not only had no difficulty in accepting it for publication, but also selected it as the best article of that year. And it was subsequently selected as one of the five AIS Best Information Systems Publications Awards for the year. There are beachheads. But evidence of those beachheads being exploited, through significant number of papers that adopt the perspectives of, in particular, users, usees and the environment, through the emergence of specialist venues, through the progressive recognition of those venues, and ultimately in maturation into a recognised instrumentalist approach, is at this stage much more difficult to find.
Senior members of the discipline need to ensure that their natural conservatism and understandable preference for a degree of stability in the discipline's scope are not so strong as to commit IS to ossification. They may be among the slowest to appreciate and declare the perspective that they are adopting in each research project; to consider adopting perspectives other than that of the system sponsor; and to respect and interact with colleagues who adopt alternative perspectives. In their gatekeeper roles, however, senior members of the discipline need to tolerate and respect diversity in the perspectives adopted by other IS researchers, and better still encourage diversity in perspectives.
To the extent that alternative perspectives are recognised as being within-scope of the IS discipline, a significant increase in the value of the academic IS literature to IS professionals can be anticipated. The first source of benefits would be an increase in the quality of research through the inclusion of all relevant factors in the design process, leading to smoother implementation, more rapid adoption, and lower system costs.
A further factor would be an increased emphasis on instrumentalist research that has direct relevance to practitioners' needs, and a proportional reduction in other forms of research which have in some cases less direct benefits, and in other cases not a great deal of practical benefit in any case.
Further, IS research would at last make material contributions to the formation and implementation of public policy. The social and environmental dimensions complement the economic, and hence IS would be contributing to holist solutions rather than only serving the needs of individual organisations and strategic partnerships.
The IS discipline is at serious risk of continuing on an inwardly-oriented spiral towards angel-on-pinhead research, producing too many papers that are of intellectual or methodological interest to a few within the discipline and a few in cognate disciplines, but not to real people. The IS discipline needs to address a whole flotilla of weaknesses.
This paper has examined the notion of the perspective from which a researcher views the phenomena of interest. It has found that researchers generally adopt a single perspective, but do so without making clear what that perspective is, even to themselves let alone to the readers of the resulting papers. This largely implicit perspective strongly influences the conception of the research, the expression of the research questions, the research design, the analysis of the evidence, and the results.
The author's strong impression that one perspective dominates IS research was examined, and found to be empirically supported. The large majority of a sample of 250 articles published in 2014 was strongly oriented towards the interests of the system-sponsor. Alternative perspectives were identified along three dimensions. On the economic dimension, the perspectives of other levels of abstraction are sometimes adopted, particularly those of joint ventures and other forms of strategic partnership. Papers that adopted the perspectives of points on the social dimension, such as users, usees, communities and societies, were uncommon, representing only 4% of the sample. Not one paper in the sample fell on the environment dimension, at any of the possible levels of abstraction such as the biosphere (the presumed ultimate beneficiary of 'green' IT activities) and the troposphere (where climate change is occurring).
Consideration was given to the question of the extent to which IS research projects reflect multiple perspectives rather than just one. This is a valuable but challenging approach, and only a single paper in the sample did so. Examples do exist in the literature, however, including a recent prize-winner. This represents a potential breach in the wall of research driven by the interests of system-sponsors.
Research approaches that extend beyond the descriptive, explanatory and predictive were reviewed. The constructivist technique of design science has to date mostly been tied to the system-sponsor perspective, whereas action research has an association with interpretivism and is much more open to multiple perspectives. The broader field of problem-driven, instrumentalist research was depicted as a step beyond both conventional interpretivist approaches, in that it inherently recognises the existence of multiple interests, begins with a problem, and seeks solutions that balance them. Critical theory research was noted as being explicitly concerned with power-structures and with altering pre-existing balances. The progressive extension of IS research into normative mode was then combined with the recognition of the social and environmental dimensions. This led naturally to the recognition that public policy research is a natural part of the scope of the discipline.
The paper concludes that the dominance of the system-sponsor perspective has resulted in lower quality than could otherwise be achieved, and is causing the discipline to miss opportunities for contributions that would be in some cases of an economic nature, but particularly in the social and environmental areas. Members of the IS discipline need to discover perspective as an important element of research conduct, deliberate on the alternative perspectives, determine the perspective(s) to be adopted, and declare the perspective(s) as part of their reports on the research. Senior members of the discipline need to ensure that they do not confuse their role as gatekeeper with the notion of palace guard, and that they permit entry to contributions that satisfy quality tests appropriate to the subject-matter.
Agarwal N., Lim M. & Wigand R. (2012) 'Raising and Rising Voices in Social Media: A Novel Methodological Approach in Studying Cyber-Collective Movements' Business & Information Systems Engineering (BISE) 4, 3 (2012) 113-126. It appeared simultaneously in German in Wirtschaftsinformatik 54, 3 (2012) 107-122
Benbasat I. & Zmud R.W. (2003) 'The Identity Crisis Within the IS Discipline: Defining and communicating the discipline's core properties' MIS Quarterly 27, 2 (2003) 183-194
Cameron J. & Clarke R. (1996) 'Towards a Theoretical Framework for Collaborative Electronic Commerce Projects involving Small and Medium-Sized Enterprises' Proc. 9th Int'l Conf. EDI-IOS, Bled, Slovenia, 10-12 June 1996, PrePrint at http://www.rogerclarke.com/EC/CamCla960612.html
Clarke R. (1988) 'Economic, Legal and Social Implications of Information Technology' MIS Qtly 12,4 (December 1988) 517-9, at http://www.rogerclarke.com/DV/ELSIC.html
Clarke R. (1990) 'Information Systems: The Scope of the Domain' Xamax Consultancy Pty Ltd, 1990, at http://www.rogerclarke.com/SOS/ISDefn.html
Clarke R. (1992) 'Extra-Organisational Systems: A Challenge to the Software Engineering Paradigm' Proc. IFIP World Congress, Madrid, September 1992, PrePrint at http://www.rogerclarke.com/SOS/PaperExtraOrgSys.html
Clarke R. (1994) 'Human Identification in Information Systems: Management Challenges and Public Policy Issues' Information Technology & People 7,4 (December 1994) 6-37, PrePrint at http://www.rogerclarke.com/DV/HumanID.html
Clarke R. (1997) 'Instrumentalist Futurism: A Tool for Examining I.T. Impacts and Implications' Xamax Consultancy Pty Ltd, October 1997, at http://www.anu.edu.au/people/Roger.Clarke/DV/InstFut.html
Clarke R. (2000) 'Appropriate Research Methods for Electronic Commerce' Xamax Consultancy Pty Ltd, April 2000, at http://www.rogerclarke.com/EC/ResMeth.html
Clarke R. (2001) 'If e-Business is Different, Then Research in e-Business is Too' Keynote Address, Proc. IFIP TC8 Working Conference on E-Commerce/E-Business, Salzburg, 22-23 June 2001, at http://www.rogerclarke.com/EC/EBR0106.html
Clarke R. (2008) 'A Retrospective on the Information Systems Discipline in Australia' Chapter 2 of Gable G.G. et al. (2008) 'The Information Systems Discipline in Australia' ANU e-Press, 2008, pp. 47-107, at http://epress.anu.edu.au/info_systems_aus/pdf_instructions.html, PrePrint at http://www.rogerclarke.com/SOS/AISHist.html
Clarke R. (2014) 'The Prospects for Consumer-Oriented Social Media' Proc. Bled eConference, June 2014, PrePrint at http://www.rogerclarke.com/II/COSM-1402.html
Clarke R. (2015) 'Big Data Quality: An Investigation using Quasi-Empirical Scenario Analysis' Proc. 28th Bled eConference, June 2015, at http://www.rogerclarke.com/EC/BDSA.html
Emery F.E. & Trist E.L. (1960) 'Socio-technical Systems' Management Sciences Models and Techniques, vol. 2, 1960
Goldkuhl G. (2012) 'Pragmatism vs interpretivism in qualitative information systems research' European Journal of Information Systems' 21, 2 (2012) 135-146, at http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-76528
Hevner A.R., March S.T. & Park S. (2004) 'Design Science in Information Systems Research' MIS Quarterly 28, 1 (March 2004) 75-105
Hirschheim R. & Klein H.K. (2012) 'A Glorious and Not-So-Short History of the Information Systems Field' Journal of the Association for Information Systems 13, 4 (2012)
Iivari, J., "A Paradigmatic Analysis of Information Systems as a Design Science, Scandinavian Journal of Information Systems, 19(2), 2007, at http://www.researchgate.net/profile/Juhani_Iivari/publication/201168950_A_paradigmatic_analysis_of_information_systems_as_a_design_science/links/0912f5062852f50bf2000000.pdf
Introna L.D. (1996) 'Notes on Ateleological Information Systems Development' Information Technology and People 9, 4 (1996) 20-39, at https://sites.google.com/site/lucasintrona/home/journal-publications/publication-archive-1-1/NotesonAteleologicaldevelopment.pdf
Keller S. & Coulthard D. (2013) ' Charting diversity and change in IS publications: A tri-continental journal analysis' Australasian Journal of Information Systems 18, 1 (2013), at http://journal.acs.org.au/index.php/ajis/article/view/832
Klein H. & Myers, M. (1999) 'A set of principles for conducting and evaluating interpretive field studies in information systems' MIS Quarterly 23, 1 (1999) 67-93, at http://www.csisdmz.ul.ie/staff/brianFit/phd-seminar-series/myers-klein-misq.pdf
Koh C., Ang S. & Straub D.W. (2004) 'IT Outsourcing Success: A Psychological Contract Perspective' Information Systems Research 15, 4 (December 2004) 356-373
Land F. & Hirschheim R. (1983) 'Participative systems design: Rationale, tools and techniques' Journal of Applied Systems Analysis, 1983
McGrath K. (2005) 'Doing critical research in information systems: a case of theory and practice not informing each other' Information Systems Journal 15, 2 (April 2005) 85-101, at http://v-scheiner.brunel.ac.uk/bitstream/2438/3734/1/Fulltext.pdf
Mumford E. (2000) 'A Socio-Technical Approach to Systems Design' Requirements Engineering 5, 2 (September 2000) 125-133
Orlikowski W.J. & Baroudi J.J. (1991) 'Studying Information Technology in Organizations: Research Approaches and Assumptions' Information Systems Research 2 : 1 (1991) 1-28
Walsham G. (1995) 'The Emergence of Interpretivism in IS Research', Information Systems Research, 6, 4 (1995) 376-394
Walsham G. (2012) 'Are we making a better world with ICTs? Reflections on a future agenda for the IS field' Journal of Information Technology 27, 2 (2012) 87-93
Roger Clarke is Principal of Xamax Consultancy Pty Ltd, Canberra. He is also a Visiting Professor in Cyberspace Law & Policy at the University of N.S.W., and a Visiting Professor in the Computer Science at the Australian National University.
The content and infrastructure for these community service pages are provided by Roger Clarke through his consultancy company, Xamax.
From the site's beginnings in August 1994 until February 2009, the infrastructure was provided by the Australian National University. During that time, the site accumulated close to 30 million hits. It passed 50 million in early 2015.
Sponsored by Bunhybee Grasslands, the extended Clarke Family, Knights of the Spatchcock and their drummer
Xamax Consultancy Pty Ltd
ACN: 002 360 456
78 Sidaway St, Chapman ACT 2611 AUSTRALIA
Tel: +61 2 6288 6916
Created: 17 April 2015 - Last Amended: 30 September 2015 by Roger Clarke - Site Last Verified: 15 February 2009
This document is at www.rogerclarke.com/SOS/EPinIS.html