A Socio-Technical Framework for Supporting Learning-Based Decision Making

As evident from the review of the three learning modes, the nature of feedback is shaped in relation to the problem-driven approach to learning. Equally it is noticeable that many of the decision-making models are also problem-driven. An example would be Simon's (1960) behavioral model, which is based on the principle of 'bounded rationality' and assumes that decision makers try to be rational, but their rationality is bounded partly by their representation of the decision problem (Barth lemy, Bisdorff, & Coppin, 2002). Sprague and Carlson (1982) have enhanced Simon's model by adding a previously implied 'Implementation' step. This is the final stage of the decision process where the decision makers implement their chosen alternative by taking action, and monitoring performance and effectiveness. If there is a mismatch between reality and a desired state, and the outcome of the action taken does not meet expectations, then the decision process or a part of it is repeated.

Therefore, the relationship between learning and decision-making reflects that they share a common structure in the way both assume the process of learning and decision-making takes place. This structure tends to be seen predominantly as linear underpinned by feedback mechanisms, which provide a cyclical approach to the way learning and decision-making unfold. The decision-learning model discussed in the previous section however, also shows that feedback mechanisms are dynamic and complex and not simply technological tools (in the various DSS solutions) which act as information depositories. In developing a socio-technical framework of learning-supported decision-making, it is important to capture the dynamic nature of feedback systems reflecting both the way ICT systems are part of the social structure which allows information to progress from meaningless data to meaningful knowledge and from knowledge to insightful learning. We seek to capture these points in the three levels of feedback, which we argue underpin the socio-technical framework of decision-learning. Figure 2 presents the three levels of feedback diagrammatically.

Figure 2: A Socio-Technical Framework for Supporting Learning-Based Decision Making

First-Level Feedback

This first level of feedback reflects a mode of decision-making geared towards implementing a solution to a perceived problem supported predominantly by single-loop learning. By virtue of this rather basic process of decision learning, the role of feedback systems is to extract, process and disseminate basic available information to develop an understanding of the perceived problem so that the best possible solution can be reached. While this level of feedback may appear rather simplistic in essence it may well be that the nature of the issue at hand requires only a basic processing of the information at hand. The typical ICT systems that can be used to facilitate this process include data marts, data warehouses, knowledge repositories, OLAP products and executive information systems. Such decision support systems assemble flows of knowledge from different sources and assist decision makers in making sense of the relevance and significance of information to a particular decision problem. Essentially, these ICT systems would be part and parcel of the existing information structure which would reduce decision times and costs to detect and correct problems in a routine setting. Their capability to extract, process and disseminate information is vital in facilitating learning in organizations (Harvey, Palmer, & Speier, 1998) and can be further enhanced using artificial intelligence techniques (Bhatt & Zaveri, 2002).

Second-Level Feedback

The mode of decision making at the second level of feedback extends the focus well beyond the decision itself and seeks to explore the decision-making process itself. This implies that decision-makers are concerned to improve not only the decisions they make but the way they take their decisions. Essentially therefore, the focus is to critically reflect on the implicit assumptions that underpin the current decision-making structures and through this process seek to develop new understanding. The double-loop learning process that underpins the focus of decision-making provides a useful feedback platform to translate available information into valuable knowledge which feeds back to current decision-making practices critical insights into the values and perceptions which underpin them. Fundamentally, second level feedback to some extent disrupts social structures that support decision-making and learning by bringing to the forth greater awareness of the actual process of learning and taking decisions. This very approach of looking critically at the way a process takes place can be greatly supported by ICT systems which have the capacity to overcome the inherent subjectivity embedded in learning and decision practices. For example, collaborative support tools (Maybury, 2001) can be used to facilitate interactions with expert decision makers who can provide insight into the content, context and process of decision making. Architectures of decision support systems that codify problem solving strategies in new entities called model marts and model warehouses have recently been introduced (Bolloju et al., 2002). Emerging technologies such as virtual reality could be used to represent experiential rather than formal forms of knowledge (Beynon, Rasmequan, & Russ, 2002). Therefore, second-level feedback encodes rich experiences and expertise, improves the efficiency of decision-making, and contributes to identity formation by reflecting on the meanings decisions made create and not just on the meanings that shape decision-making.

Third-Level Feedback

The third level of feedback, essentially embraces learning as central to both the decision-making and learning process. In other words, it extends the focus of decision-making and learning on the people performing this process by virtue of their actions not just assumptions. Consequently, the focus of decision-making would be reflecting on the process of learning to take decisions rather than the decision-making process itself. This also would extend to the approach to learning which would also be concerned with learning how to learn (Bateson, 1972). The feedback systems therefore, would be concerned with reflecting the social structures that support learning to learn using ICT systems that provide mechanisms for capturing and supporting the learning process. Even though ICTs that enhance discourse and communication can facilitate learning (Robey, Boudreau, & Rose, 2000), they cannot always substitute the importance of face-to-face meetings and informal communication. However, decision practices have changed in the recent years and organizations are now forced to support geographically dispersed teams of decision makers. ICTs such as videoconferencing and group support systems allow the exchange of social cues—in the case of text-based tools the cues are known as 'smileys' or 'emoticons'—and in doing so facilitate and enrich communication (Shim et al., 2002). Such ICTs therefore, could be embedded in the organizational discourse and facilitate communicational exchanges by addressing linguistic interpretations, cognitive representations and reciprocal learning (Barth lemy et al., 2002). The main characteristics of the three feedback levels are summarized in Table 2.

Table 2: The Characteristics of Three Levels of Feedback

Feedback level	First Level	Second Level	Third Level
Source of feedback	Knowledge repositories, search engines	Lessons-learned systems, experts	Communities of practice
Type of feedback	Passive	Active	Organic
Attitude of decision maker	Reactive	Reactive/Proactive	Proactive
View of decision making	Atomistic	Group	Social
Focus	Decision content, context, process	Values and assumptions	Decision-making capabilities
Tactic	Make knowledge easily retrievable	Make meta-knowledge and experts easily accessible	Set up and nurture communities of practice
Purpose	Improve decision-making efficiency and effectiveness	Improve decision quality	Improve decision-making capabilities

Therefore, it would be idealistic to suggest that learning-supported decision-making would be reflected through deutero learning, supported by deutero decision-learning (in the use of ICTs) in an approach that facilitates decision-making based on feedback level three. The socio-technical framework of decision-making needs to be based on a balanced approach between social systems and technical structures so that different feedback levels can be effectively integrated to support different stages in the decision-making and learning process. ICT systems can form the backbone of different levels of feedback in capturing the dominant discourse and its associated social structures. This point suggests that the socio-technical framework relies on the homogeneity within the heterogeneity of decision-makers which is never settled or fixed but constantly in progress as multiple perspectives given and perspectives taken are negotiated as part of the emerging organizational discourse.