Firstly I would like to thank the organisers for inviting me to speak at this conference to present my paper to you. The paper that I am presenting to you today is titled, ‘The role of mediating artefacts in the design of persuasive e-learning systems’, in which I have presented a model for designing Internet applications that persuade users to carry out specific actions in order to meet their learning goals.
You’re probably wondering who I am and what my background is. My name is Jonathan Bishop, it was just over a year ago that I graduated from the University of Glamorgan with a Masters of Science in E-Learning, after completing an honours degree and HND in Multimedia. Glamorgan and this establishment both have something in common – they both run excellent multimedia courses and they both appeared on my UCAS form over six years ago. I could be speaking to you today as a graduate of this fine institution, but whilst I did not study here, I am pleased to have been given the opportunity to present my paper to you today.
It was on the Masters in E-Learning at Glamorgan that I started to think about the role of mediating artefacts in the design of persuasive e-learning systems and my thesis about educating heterogeneous user groups using persuasive technology touched upon the topic.
I have borrowed the term mediating artefact from the work of L. Vygotsky, who saw artefacts as mediators for action. An artefact can be more clearly thought of as a tool or a sign. Those who have built upon Vygotsky’s work see the computer as a mediating artefact; they see it as a tool for carrying out actions. I believe this is the wrong interpretation to take of computers. According to my model of ecological cognition, the computer is not an artefact, but an environment that contains artefacts. According to my model, a mediating artefact is a tool, sign or other object in the environment that offers an affordance. For example, in the real world a door handle can offer the affordance of pulling, and in the virtual world a graphic can offer the perceived affordance of clicking.
In a virtual environment, such as an Internet application, a mediating artefact can be seen as any object to which an event handler can be assigned. For example a hyperlink can have the OnClick event handler assigned to it and an icon can have the OnDrag event handler assigned to it. All artefacts with event handlers on them can be considered to be mediating artefacts as they exist to facilitate action. Those artefacts that make their purpose and function clear to the user and therefore offer an affordance can be considered to be persuasive mediating artefacts, as they encourage the user to take an action, such as clicking or dragging.
Internet applications, including e-learning systems are made up of mediating artefacts, some of which are persuasive and offer affordances and others that do not make their purpose explicit.
The purpose of my paper was to develop a model that would allow developers of e-learning systems to design applications that persuade users to use mediating artefacts in ways that would allow them to meet their goals. The model I developed suggests four stages to developing persuasive e-learning systems; the first stage is to identify the user’s goals, the second stage is to identify the information users need to achieve their goals, the third stage is to identify the mediating artefacts users need to achieve their goals, and the fourth and final stage is to develop the system and evaluate its persuasiveness.
The first stage of the model is to develop the scenarios of the existing system, where one exists, to identify the goals of the users in order to design a new system. Understanding how the existing system works forms a core part of many system analysis and design methodologies, which suggests that any new methodology should incorporate it.
I propose using scenarios as the basis for understanding the existing system, as these are very flexible and easy to generate. These scenarios can be used to identify the goals of the system users and the context in which they carry out actions in order to meet them.
J.M. Carroll has suggested using scenarios in several publications since 1994 as an effective means of conceptualising what users want to achieve with a computer application, such as an e-learning system. A scenario suggests the goals of a user, depicting a situation in which they carrying out an activity and explains what they are trying to achieve. In my paper I used the example of Edward, a person interested in astronomy who is looking for information on black holes and the scenarios explains what he is doing in order to meet his goals.
The second stage of the model is to identify the information required by the user to meet their goals. Using the example of Edward, who has a goal to learn about astronomy and in particular black holes we can see that in order to meet this goal, Edward needs to know where information on astronomy is located, so this is the first information requirement. Edward also had a goal to store information for later use, making the second information requirement knowing where this information would be stored.
The third stage of the model is to identify the mediating artefacts that user needs to achieve their goals. To meet Edward’s goal of locating material on black holes, we could provide mediating artefacts such as text or graphics that offer the perceived affordance of clicking. My study showed that mediating artefacts that conveyed an action, such as a button with ‘search’ on it or a graphic with a functional symbol on it persuaded users of all abilities to carry out actions to meet their goals.
The fourth stage of the model is to develop the system and evaluate its persuasiveness.
There are many suitable methods for developing e-learning systems, such as rapid prototyping and instructional design methods. However, less popular methods, such as paper prototyping and storyboarding can be more effective when it comes to deciding how to integrate the mediating artefacts identified at stage 3 into an e-learning system. These methodologies allow the interaction designer to try out different ideas for the location of mediating artefacts in a cheap and affordable way, meaning they can work out the most effective way of persuading the user to use specific artefacts to achieve their goals.
Once the system has been developed, the persuasiveness of the mediating artefacts and the system as a whole need to be evaluated. Whilst there has been a lot of research that has identified the most effective ways of evaluating the usability of an e-learning system, there is little research on methods to establish the persuasiveness of a system. A study by P.K. Murphy and others compared the persuasiveness of online and printed material used a methodology that tested the knowledge and beliefs of participants before and after being exposed to the persuasive material. This method could be successful at establishing whether the content of the e-learning system is persuasive. However, the difficulty with this method is that it only tests the overall persuasiveness of the system and not the effectiveness of the persuasive mediating artefacts.
In the study I carried out I used the Wizard of Oz technique to evaluate the persuasiveness of the mediating artefacts used in the system. This technique involves the evaluator changing aspects of the system and monitoring how the user responds to them. I found that users of all levels of experience respond quickly to the graphical mediating artefacts that were explicit in their purpose, such as the graphic that displayed a ‘play’ button, and also found users clicked on buttons that suggest an action, such as a button with ‘search’ on it.