MIS Doctoral Seminar

In recent years, more researchers and practitioners realize the uses and consequences of information technology are often enacted through self-orchestrated interactions among users, technologies, and institutional properties rather than dictated by organizational policies or managerial intentions. To further investigate the pattern, the paper aims to build a theoretical framework based on the concepts and the analytical tool of complex adaptive systems (CAS) theory.

Conceptualization of IT Use

There are four essential attributes of the IT use construct. IT use involves users, IT artifacts and work activities as its key elements. The behaviors and consequences are shaped by recurrent interactions within and between different elements of IT use. IT use processes operate at multiple levels linked by bottom-up or top-down processes. Time is a significant factor.

A Complex Adaptive Systems Model of IT Use

Theory of complex adaptive systems

Complex adaptive systems are defined as “systems composed of interacting agents described in terms of rules. The agents adapt by changing their rules as experience accumulates”. Three are three main components in the framework: agent, interaction and environment. They also have their own properties.

Extending the theory of CAS to the IT use context                                               

The CAS model of IT use is premised on the multilevel, interactive, and dynamic nature of the IT use construct. It is developed to provide a theoretical lens for examining the process whereby individual-level IT use behaviors collaboratively create collective-level IT use patterns and outcomes.   

• Agent: human actors, IT features
• Interaction: user-system interactions, interpersonal interactions
• Environment

  Using the CAS Model of IT Use       

Based on the Zeta case and the CAS framework, the author designs the agent-based modeling by specifying the conceptual definition for each element, including the roles and learning process of Zeta employees,  the ITSS (incident tracking support system) and its feature, interactions between users and users, interactions between users and the system, requirements of the customer support work  and the IT-based transformation.

Discussion

• Implications for theory: The CAS model can offer advancements of the three dominant theories of IT use
• Implications for methodology:  Agent-based modeling contributes to interpretive case studies by providing an analytically precise and consistent method for describing and comparing qualitative observations from the field.

 

After holistic thinking was institutionalized in the 1950s, the concept of ‘system’ and ‘systems thinking’ have been explicitly used. In 1954, general systems theory (GST), the development of a mathematically expressed general theory of system, was supposed to provide a meta-level language and theory in which the problems of many different disciplines could be expressed and solved. GST, however, has not succeeded.

This paper provided an example of the third kind of development: the combinations of the systems ideas and the use of systems ideas.

Hard and soft systems stances

Hard systems stance assumes that the world can be taken to be a set of interacting systems, some of which do not work very well and can be engineered to work better.

Soft systems stance assumes that the world is taken to be very complex, problematical, mysterious. However, our coping with it, the process of inquiry into it, it is assumed, can itself be organized as a learning system.

Principles of SSM

1. Perceived a real-world problem situation

• Some ways for the grasp of the problem situation as rich as possible
• Pictures are a better medium than linear prose for expressing complex relationships.

 

2. Relationships explored via models of purposeful activity based on explicit world-views

• The purposeful activity models used in SSM are intellectual devices whose role is to help structure an exploration of the problem situation being addressed.

 

3. Inquiry structured by questioning perceived situation using the models as a source of questions

 

4. ‘Action to improve’ based on finding accommodations (versions of the situation which conflicting interests can live with)

 

5. Inquiry in principle never-ending; best conducted with wide range of interested parties; give the process away to people in the situation

 

Conclusions

Learning SSM is to learn not what to think, but a way of thinking. The biggest difficulty in understanding SSM is to absorb its shift from assuming ‘systems’ exist in the world (as in everyday language) to assuming that the process of inquiry into the world can be a consciously organized learning system.

This is a summary of the work Checkland, P. (2000). Soft systems methodology: a thirty year retrospective. Systems research and behavioral science, 17(S1), S11-S58.

The holistic thinking of a newly created organization, that which makes explicit use of the concept of ‘system’, and ‘systems thinking’ is essential. This article provides an example of the third kind of development by combining two development: the work of Maturana and Varela (1980) and Chorley and Kennedy (1971). Rather than make neither the ideas nor the practical experience dominant, this work’s intention was to allow the tentative ideas to inform the practice which then became the source of enriched ideas – and so on, round a learning cycle. This is called the action research cycle whose emergence of action research cycle is described in Systems Thinking, Systems Practice and whose use and development is the subject of SSM in Action.

Soft System Methodology (SSM)

The action research programme is to find ways of understanding and coping with the perplexing difficulties of taking action, both individually and in groups, to ‘improve’ the situations which day-to-day life continuously creates and continually changes

The intellectual starting point was that an organization could be taken to be a system with functional sub-systems. Given such an explicit definition of an objective, a system can in principle be ‘engineered’ to achieve that end. Then we found that although we were armed with the methodology of systems engineering and were eager to use its techniques to achieve the objectives, the management situations were always too complex for a straightforward application of the systems engineering approach.

Four key thoughts of the development of SSM

1. Modelling purposeful ‘human activity systems’ as sets of linked activities which together could exhibit the emergent property of purposefulness.
2. Select the human activities which are likely to be most relevant in exploring the situation;
3. Declare a world-view which made a chosen model relevant and would determine the model content;
4. Understanding the distinction of ‘hard’ system and ‘soft’ system.

Seven Stages model of SSM

The first two stages entail entering the problem situation, finding out about it and expressing its nature; Stage three is root definition; Stage four is modelling. The next stages use the models to structure further questioning of the situation (stage five ‘comparison’). Stage six is to seek to define the changes which could improve the situation. Stage Seven is the process of implementing a real-world change.

Two streams of SSM

SSM is an approach embodying not only a logic-based stream of analysis but also a cultural and political stream.

Prescriptive and Internalized SSM

Prescriptive SSM Model	Internalized SSM Model
Methodology-driven	situation-driven
Intervention	interaction
Sometimes sequential	always iterative
SSM an external recipe	SSM an internalized model

Previous research about system usage mainly focus on the patterns and outcomes of collective-level information technology use, however, in this paper, it explored  the key properties and mechanism of bottom-up IT use process

IT use

• Conceptualization: essential attributes
  • Key elements: users, IT features, and tasks
  • Recurrent interactions within and between different elements of IT use => ( IT use behaviors + consequences )
  • Operate at multiple levels linked by bottom-up and top-down processes
  • Time is a significant factor in patterns and outcomes of IT use.
• Literature
  • Individual-level IT use
  • Collective-level IT use
  • Interactions among the key elements of IT use
  • Dynamic patterns of IT use
  • The top-down process between the individual and collective levels of IT use (rare)
  • The bottom-up process between the individual and collective levels of IT use

Complex Adaptive Systems Model

• Conceptualization: three components
  • Agent: individual actors or basic entities of actors
    • Attributes: internal states of agents (change over time)
    • Behavioral rules: schema governing attributes and behaviors
  • Interaction: mutually adaptive behaviors of agents
    • Agents: processors of interactions
    • Connections: relational links among agents
    • Flows: movements of resources through the web of agents and connections
  • Environment: medium for agents to operate in and interact with
  • Agents move toward resource-rich locations in an environment
• Extension to IT use
  • Agent: human actors & IT features
  • Interaction: user-system interactions & interpersonal interactions
  • Environment: social or organizational contexts of IT use processes

A case study of CAS: Zeta

• Agent:
  • Human actor -> employees (learning rate)
  • IT features -> ITSS features (flexibility or the ease of modifying its features)
• Interactions:
  • user-system: direct uses of the ITSS
  • interpersonal: interpersonal ties causing mutual-influence employee behaviors
• Environment:
  • Organizational context
  • Structure: organizational policies, culture

Simulation: agent-based model

• Higher learning rates may impede work performance gains
• Higher ITSS flexibility may impede ITSS assimilation and work performance gains
• Rigid workplace may facilitate ITSS assimilation and work performance gains

This article presents a systematic approach for reconceptualizing the system usage construct in particular nomological contexts. The approach comprises two stages: definition and selection.

Two concerns that need to reconceptualize system usage: no theory and poor-to-no validation

A staged approach for reconceptualizing system usage:

A two-step method for selecting measures of usage in future research: system usage can be attributed with a precise definition, but the definition refers to a broad range of content, only a subset of which will be relevant in a specific study.

Define stage:

Define the distinguishing characteristics of system usage and state assumptions regarding these characteristics.

==> Selection stage:

Choose the best measures for the part of the usage activity that is of interest.                           

Step 1: Structure. Select the elements of usage that are most relevant for the research model and context.

Step 2: Function. Select measures for the chosen elements that tie to the other constructs in the nomological network.

The staged approach advanced in this paper helps to clarify the meaning of system usage and the range and dimensionality of past usage measures.

 

Research Question

How could we reconceptualize the system usage construct to provide a strong theoretical basis?

Research Objectives

This paper presents a systematic approach for reconceptualizing the system usage construct in particular contexts. The approach consists of two stages, definition and selection, allow researchers to solve issues of the lack of theory and lack of validation in system usage studies.

Why We Need to Reconceptualize System Usage?

Two concerns: 1) No theory. Few studies discuss how theory informs their choice of usage measures. In absence of a theoretical basis, an abundance of measures could be problematical. 2) No validation. Almost no validation of the usage construct has been conducted. After all, past studies have provided mixed conclusions about the link between system usage and individual performance.

A Staged Approach for Reconceptualizing System Usage

1. Definition Stage: Define the distinguishing characteristics of system usage and state assumptions regarding these characteristics.
2. Selection Stage: Choose the best measures for the part of the usage activity that is of interest.

• Step 1: Structure. Select the elements of usage that are most relevant for the research model and context.
• Step 2: Function. Select measures for the chosen elements that tie to the other constructs in the nomological network.

Empirical Investigations of the Approach

Using the two-stage approach, authors develop a model of system usage and individual task performance. The result from the approach is a higher-order construct, exploitive system usage, which is expected to be more beneficial than other less strong constructs such as cognitive absorption and deep structure usage.

Empirical Test of the Approach

To compare performances of measures, authors conduct a lab experiment to university students. The results show that a very rich measure of usage (exploitive system usage) yields almost three times the variance explained by the lean measure (duration) and consistent positive effects on performance.

Contributions

This paper provides a way to explicitly conceptualize and reconceptualize system usage in IS research.

Previous research about system usage on different level

• Different levels:
  • Individual: behavior, cognition and affect
  • Group: an aggregation of individual level behavior, a pattern of individual behaviors, cognitions and affect
  • Organization: an aggregation of individual level behavior, an intra- (and/or inter-) organizational behavior
• Definition of usage: user (subject) + system (IS) + task (performed)

Motivations for multilevel research

• Avoid significant fallacies (cross-level, contextual, ecological and atomistic fallacy) that can occur in single-level studies
• Open up new opportunities for theory and may even help generate new organization-specific theories

Nature of multilevel constructs: application of system usage

• Function of usage:
  • Individual level: use of an IS influences user performance
  • Collective level: users actively reflect on the results of their usage and adjust their actions over time
• Structure of usage:
  • Interaction (interdependencies-in-use)
    • Collective usage exist: strength of interdependencies-in-use are moderate or strong no matter the proportion of members interacting with the IS
    • Guidance: be mediated in several ways; be sure to measure them; whether and why they exist in a given scenario.
    • Three types: collaboration, communication and coordination
  • Development process (form of collective use)
    • Global: level of origin == level of theory
    • Shared: (level of origin < level of theory) & form of homogeneous pattern
    • Configural: (level of origin < level of theory) & form of distinct pattern
  • Context (context of usage)
    • Context-function: contextual factor associated with each element of usage; given structure of collective usage associated with different outcomes due to differences in time-scale
    • Context-structure: nature of each element of usage can facilitate construction of interaction; time is an important contextual factor with collective properties

This article concerns theories about why and how information technology affects organizational life, and examines theories in terms of their structures. Three dimensions of casual structure are considered:

Causal Agent

The analyst’s beliefs about the identity of the causal agent, the nature of causal action and the direction of causal influence among the elements in a theory.

Three conceptions of causal agency in the literature on information technology and organizational change:

• The technological imperative:

It views technology as an exogenous force which determines or strongly constrains the behavior of individuals and organizations.

• The organizational imperative:

It assumes almost unlimited choice over technological options and almost unlimited control over the consequences.

• The emergent perspective:

The uses and consequences of information technology emerge unpredictably from complex social interactions.

Analysts of different persuasions can often be distinguished as easily by their normative stance as by the specifics of their models.

Logical Structure

The temporal aspect of theory – static versus dynamic – and to the logical relationships between the “causes” and the outcomes.

	Variance Theory	Process Theory
Role of Time	Static	Longitudinal
Definition	The cause is necessary and sufficient for the outcome	Causation consists of necessary conditions in sequence; change and random events play a role
Assumption	The outcome will invariably occur when necessary and sufficient conditions are present	Outcomes may not occur (even when conditions are present)
Elements	Variables	Discrete outcomes
Logical Form	If X, then Y; If more X, then more Y	If not X, then not Y; cannot be extended to “more X” or “more Y”

Level of Analysis

The entities about which the theory poses concepts and relationships

• Macro-level theory: a productive strategy for generating falsifiable, parsimonious, and readily comprehensible explanations for behavior
• Micro-level theory: only people can act
• Mixed-level theory: technologies are neither strictly micro nor macro in character

 

About Theory

• The author discusses theory in terms of the nature of theory in general, generalization, causality, and goals (explanation and prediction).
• Theory, in this study, is “abstract entities that aim to describe, explain, and enhance understanding of the world and, in some cases, to provide predictions of what will happen in the future and to give a basis for intervention and action.”
• Generalization
  • Theories can be classified such as Meta-theory, grand theories, substantive theory, formal theory, and mid-range theory based on their level of generalization and breadth of focus.
• Causality
  • Four prominent approaches to the analysis of causality a) regularity (nomological) analysis (arguably most popular in natural science), b) counterfactual analysis (closely related to regularity analysis), c) probabilistic causal analysis (suited to social science), and d) manipulation or teleological causal analysis
• Explanation and prediction
  • Explanation can refer to two cases: a) a particular phenomenon can be explained through the scientific law; and b) explanation depends not on universal law, but on other causal mechanisms such as teleological causes. This study addresses the second type of explanation.
  • Other than explanation, theories may aim at prediction.

Classifying theory in IS

• A taxonomy of IS theory types includes a) analysis, b) explanation, c) prediction, d) explanation and prediction, and; e) design and action.
• Structural component of theory are a) means of representation, b) constructs, c) statements of relationship, d) scope, e) causal explanations, f) testable propositions, and g) prescriptive statements.

Five types of theory in IS

• Type I: Theory for analyzing
  • This kind of theory suggests “what is” contrary to causality or prediction.
  • It is needed when we have no or very limited understanding of the phenomena in question.
  • This type of theory includes classification, schema, frameworks, or taxonomies.
• Type II: Theory for explaining
  • The theory says what is, how, why, when, and where, but without making testable predictions.
  • Type II theory is identified at high- and low-level:: a) one as enlightenment at a high level and b) one for explaining some particular real-world situation.
  • This kind of theory should be new and interesting and explain previously poorly understood phenomena with plausibility and credibility.
• Type III: THeory for predicting
  • The theory says what is and what will be, but without a why part remaining a ‘black box’.
  • This is because a) prediction is of particular interest to theories and b) reasons to justify the empirical results have not yet been discovered, and c) a certain theoretical view (e.g., logical positivist) leans toward not including causality.
• Type IV: Theory for explaining and predicting (EP Theory)
  • The theory says what is, how, why, when, where, and what will be.
  • Example: Technology Acceptance Model (TAM) which aims to both explain and predict.
• Type V: Theory for design and action
  • The theory says how to do something with explicit prescriptions.
  • Example: Design science in IS
  • Design theory includes methodologies and prescriptions for developing specific applications.
  • The criteria for evaluating contribution of Type V research include utility to users, the novelty of an artifact, and persuasiveness.
• Interrelationships among theory types
  • Type I theory is the basis for all other theories by providing phenomena in question.
  • Type V theory, design science, can be informed by all other theories.
  • Particularly, type IV and V are closely related to each other in that knowledge informs the design of IT artifacts and such artifacts can be explained through EP theory.

Application of the taxonomy

• The authors applied the theory types to issues of MISQ and ISR.
• The results show that type IV is the most frequent and type III is the least.

Questions for discussion

• One theory type does not represent a particular paradigm. They are not clearly connected.
• Each theory type has its own values and contributions. We cannot rank them.
• There is controversy over whether each theory type can be called theory. The authors argue that labeling various forms of theory as a theory has a number of advantages such as complementarity among theory types and motivating researchers to think clearly about theory.
• One type of theory (e.g., Type I) can precede the other.
• The taxonomy proposed might be different in other fields because of IT artifacts which those fields lack.