Proposal view
Proposal Type: Individual Paper 
Domain: Learning and Instructional Technology 
SIG: Instructional Design 
Type Submitted Paper 
Equipment PC and projector
Paper Details
Title Framework: Deriving, designing, and applying dynamic stories in information security
Abstract

“If stories are powerful, then we need to understand why they are in order to benefit from their potential for knowledge transfer ...” (Swap et al. 2001, p.106).


The use of stories to support learning is not new (McEwan & Egan 1995). Indeed, stories based on the experience of experts commonly used in instruction involving complex domains to reflect an expert’s individual experience and to convey in a meaningful context recognized principles of best practices. However, such stories may not represent the relevant complexity of a particular system and they may not promote deep insight and improved performance with regard to problem solving in complex domains. The innovative approach reported herein makes use of stories that are generated from formal system models (SD), developed by the AMBASEC[1] project.  To derive dynamic stories from computer-based models guarantees that they have both, authenticity and relevance to the behavior of the modeled system. In this paper, we distinguish different kinds and parts of dynamic stories and indicate how they can be used to support a variety of learning goals and situations. The development of a framework is needed to strengthen the concept of dynamic stories for the application in Information Security[2].







[1] AMBASEC = A Model Based Approach to Security Culture




[2] Information Security = all aspects related to defining, achieving, and main­taining confidentiality, integrity, availability, non-repudiation, accountability, authenticity, and reliability of information or information processing facilities.



Summary



Aims and significance of research



The representation of complex problems via simulation environments to enable and improve understanding is a powerful use of a cognitive tool. For different reasons, the willingness or the ability of learners to gain insights using such models is sub-optimal. For this reason, we make use of the educational impact of stories (Schank, 1990). A dynamic story mirrors the complex content of a model and its behavior. These models contain crucial expert knowledge which remains hidden if not explained.



Methodology




System dynamics (SD) is a methodology that is appropriate for explaining and understanding complex systems (e.g., information security systems). We opt for the term dynamic stories to stress that they are generated from system dynamics models and that they represent stories describing behavior over time. Such models can tell potentially hundreds of stories. The question is ’what should be told from a complex model to achieve an instructional objective’? We have been using a rudimentary approach to elicit crucial content we call the ‘call-out box technique’. In addition, we are proposing a structured validation approach. The second challenge is the instructional design. The dissemination of the encapsulated expert knowledge within the model using dynamic stories needs a useful learning environment to evoke and enable intentional learning processes. The third challenge is to write scripts for dynamic stories which are defining how the parts of the dynamic stories are used for educational purposes.


Figure. 1.


Types of Dynamic Stories

A prerequisite for designing ‘instructional media’ is the determination of the training goal and the targeted group.  For this we suggest three main types of dynamics stories. Hereby we assume an unintended ‘trade off’ between the coverage of knowledge (how many learners can be reached) and the potential learning outcome (level of processing).


    Developing a database (the expert knowledge is stored in an ‘easy to read‘ mode)


    A paper and pencil learning approach which reveals situation, problem, and solution


    As one core element of a workshop (very close to case method)



Parts of Dynamic Stories

Loans are made from the ‘case method’ to develop distinct parts of dynamic stories:


·  Case study method - the problem has to be identified (i)


·  Case problem method  - the problem is explicit represented; solution has to be derived (ii)


·  Case incident method - the case is presented with missing data (iii)


·  Stated-problem method - the given problem and the solution have to be reflected (iv)


Basic elements of these distinctions especially the case study method (i) and the case-problem method (ii) and in addition Wolstenholme’s idea of problem- and solution-archetypes (2004) have been taken to distinguish three ‘parts’ of dynamic stories. These parts are interlinked.


 

Figure. 2.


Part 1: 

Part 1 of a dynamic story only describes the ‘situation’. Facts, behavior and specific data are revealed. Hereby the learner is ‘situated’ (Resnik, 1991; Brown et al., 1989) within the story as a responsible, but without any specific task. To detect the problem that is to discover additional side effects or not intended effects beside the perceived system boundary.


Part 2:

Unlike part 1 of a dynamic story, part 2 describes the situation and problem in detail. This has an affinity to ‘problem archetypes’.


Part 3: 

The whole dynamic story is presented. There is seldom just one easy ‘solution’ for a complex problem. So we do offer one possible solution to the described problem. This does not exclude that the learner may find better or alternative solutions.


There are several reasons to proceed like we suggested. Before problems are solved they first have to be detected. To improve the ability to perceive potential problems seems to be a step which comes first. In the field of information security ‘situational awareness’ plays an important role. If people do not perceive that things are going wrong, they aren’t able to react in time and appropriately.


Scripts for Dynamic Stories

The steps (script) for the application of a dynamic story has to be adjusted whether a rather self directed learning approach or a workshop approach is scheduled. This depends on the application - type of a dynamic story. This adjustment process is based on the triangle concept of dynamics stories (see figure 1.).


Findings


How to find out what the situation, the problem and the solution is in a computer-based model (SD)?


A partly explanation one can find by the SD-methodology itself. If you stop a simulation run or make a ‘snapshot-test’ like Sterman (2000, 199) calls it, the state of the system remains (levels can be seen as well as the constants). The feedback structure which is a responsible for the ‘problems’ in a system is hidden as well as the solution. One problem which emerges applying the ‘snapshot test’ is the elimination of the main flows in a model. The basic flows represent knowledge which roughly spoken represent the physical structure of a system. Using a ‘snapshot-test’ seems to be lacking in respect to the target of the identification and validation of a dynamic story and its parts. Referring to a taxonomy about categories of knowledge (Hillen, 2004) this problem can be faced better. A general distinction one can find here is between knowledge about a system and knowledge about policies to control a system. This seems more appropriate because it includes the knowledge about the flows and the like.


Validation by notation-comparison

To enable a model's simulation it has to be quantified by defining equations. These are underlaying the model and are easy to access. The equations can be translated into 'natural' language. Comparing the translated equations with the content of the story it can be stated if all data is included.



Validation by model-reconstruction

The congruency between the designed dynamic story and the original model (SD) can be validated by reconstructing a new model (SD). To make this validation approach even more reliable different modellers can be asked to construct new models to one designed dynamic story.


 

Keywords Instructional design/development
Knowledge handling
Appendices Figure1_a.jpg 
Figure2_a.jpg
Authors
Name Surname Institution Country e-mail EARLI Number Presenting
Stefanie A. Hillen Agder University College Norway stefanie.a.hillen@hia.no   *  
Jose J. Gonzalez Agder University College Norway Jose.J.Gonzalez@hia.no    
Visit NQcontent
© European Association for Research on Learning and Instruction, 2012 All rights reserved.