Computer Aided Ethical IT Systems Design

AUTHOR
Iordanis Kavathatzopoulos and Mikael Laaksoharju

ABSTRACT

Usability of IT systems, defined in terms of Human-Computer Interaction (HCI), is based on research conducted in many scientific disciplines as well as inside the frame of HCI. It is understood as knowledge on how to construct and use an IT system with respect to what is significant regarding human cognition, perception, ergonomics, group processes, organizational structures, etc.

HCI is the discipline that investigates how all this can be integrated into IT systems construction and use processes. It is also a discipline that focuses on the development of methods and tools that can be used by designers to produce knowledge on how to build usable IT systems. Guidelines, standards and recommendations cannot cover all problems or produce detailed answers to concrete design and use problems. They are general and they have to be interpreted and adapted.

Ethical aspects are important and we have to be able to consider them, too. All above remarks are also valid regarding the problem of ethical usability: Computer Ethics as a scientific discipline

  • produces knowledge that can help us in our effort to achieve ethical usability of IT systems. It can point to the significant issues and it can provide the main principles and ethical guidelines.
  • helps gathering relevant information, interpreting it and applying it in concrete design projects.
  • focuses also on development of methods and tools that produce detailed knowledge on how to design ethical IT systems, for example VSD, Paramedic, etc.

There is however a significant difference between normal usability issues and ethical usability. In ethics, ethical choice, ethical problem solving and decision making there is another dimension which is very important. Philosophy and psychology have clearly pointed to the ability of thinking in the right way, and how this can be developed, sustained and applied on moral problems (Kant, 2006; Piaget, 1932; Kohlberg, 1985).

Given now that in ethics no one can provide detailed and functioning answers, this dimension, i.e. ethical ability, is necessary to consider and to incorporate explicitly in an ethical usability process. And this is something that does not have the same value regarding other forms of IT systems usability. Simply it is not enough to have access to a body of usability knowledge or to methods producing such knowledge. In ethical usability we need also to provide support for the acquisition and use of ethical skills.

Accordingly our methods and tools stimulate the cognitive and group/organizational mechanisms of ethical competence in combination with their function of producing knowledge of how to design ethical IT systems. EthXpert, is a computerized tool based on these theoretical assumptions, (for more information see the web site of the tool http://www.it.uu.se/research/project/ethcomp/index.php?artikel=60&lang=en or see Laaksoharju & Kavathatzopoulos, in press).

The purpose with EthXpert is to help an analyst or decision maker to understand how different design solutions affect the interests of each involved stakeholder. To support this understanding, the analysis is made explicit by iterating a procedure comprising three main steps. The first step is to create an overview by drawing a stakeholder network, i.e. a map over the relations between all stakeholders. Second the impact of each stakeholder’s interests on other stakeholders are analyzed and noted. Finally the considerations for each interest are used as foundation for making assumptions about how the stakeholders are affected by different design solutions. Not only does this process help people to scrutinize, structure and get overview of an ethical problem. The resulting document can also be used as vindication of the choices that are made.

Various ethical support systems have targeted the concern of identifying relevant information in different ways. In Paramedic Ethics (Collins & Miller, 1992) focus is put on the obligations and responsibilities of the decision maker. Based on these, the user is establishing relationships between stakeholders and then identifying considerations for the different opportunities and vulnerabilities that come from alternative solutions. Finally a negotiated social contract alternative is evaluated as a possible compromise solution. In SoDIS (Gotterbarn, 2002; Szejko, 2002) the user is first gathering extensive background information about the problem and its stakeholders and is then prompted to answer questions aimed at identifying known causes for moral problems. In ETHOS (Mancherjee & Sodan, 2004) the user is advocated to identify the open moral questions at hand through taking the role of a moral agent after which the utility of alternative solutions are quantified according to ethical theories.

It should be noted that the first two of these systems are intended for computer professionals working in technical development projects while ETHOS, like EthXpert, is not targeting a specific audience and does not assume any specific content in the problem to be analyzed. This wide application scope makes it impossible to guide the user by asking questions about previously known sources for moral problems and other means to raise awareness of ethical issues need to be deployed. We in fact consider this absence of framework for issue identification as strength when it comes to widening the agenda for the problem situation.

Further, EthXpert’s omission of imposed comparison to ethical theory, as is the case in ETHOS, forces the user to make an independent decision about the correctness of the outcome. The user is thus never lured into the false comfort in believing that a premature analysis is finished. Following the definition of autonomy, the user has to independently decide when an analysis is finished. Such a setup enforces that the responsibility for a satisfactory analysis rests with the user. Our approach has many similarities to other ethical computerized tools suggested previously, or to other ethical usability tools such as Value Sensitive Design proposed by Friedman, Kahn and Borning (2008). However, those approaches do not focus exclusively on what psychological theory and research describe as the basis of competent ethical problem solving and decision making, namely the tension between heteronomous and autonomous moral reasoning (Kohlberg, 1985; Piaget, 1932). Following that what we need are tools that promote autonomy and hinder heteronomy. All above tools are excellent to systematize, organize and take control of designer’s thinking on concrete ethical usability issues. Nevertheless, since these tools, in different degrees, urge and lead the extension of thinking to moral philosophical considerations and other details there is a risk of being too complex and of missing the main goal, namely blocking heteronomous thinking. Focus should be on how to handle practical problems. Of course that may be also an effect of Paramedic, SoDIS, ETHOS or VSD but they include analysis of or comparison between different normative moral theories, or some others are even built to propose moral solutions (for example Davidrajuh, 2008). Ethical autonomy is not at focus there nor is it considered explicitly, meaning that the control of this necessary ethical problem-solving and decision-making process is not secured.

EthXpert has been applied on the design of different IT systems with very positive results. With help from EthXpert the test groups were able to extend previous analyses through identifying additional stakeholders and interests. The procedure also gave insight in how the interests of different stakeholders were interrelated. Some of the test groups especially appreciated the collaboration feature of EthXpert. An ethical analysis often brings up many big and small issues to consider and it is therefore efficient if a group can cooperate in solving the problems. Thus the tool also works as a means to gather several perspectives on a problem.

Through the explicit process, the designer acquires both a better overview of the complexity of a problem and a conception of how the involved stakeholders affect and are affected by different solutions. Almost all of the test subjects were of the opinion that the systematic procedure of EthXpert is purposeful for acquiring higher ethical problem-solving and decision-making skills by offering a holistic overview over ethical aspects in the design of IT systems. Although critical remarks about the usability of the interface, many also became aware of shortages in a prior analysis made without the tool. This indicates that a computerized tool that guides the investigation of stakeholders’ interests, and supports structuring and overview over information, is helpful for designing more ethical IT systems.

REFERENCES

Collins, W. R. and Miller, K. W.: 1992, ‘Paramedic ethics for computer professionals’. Journal of Systems Software 17, 23-38.

Davidrajuh, R.: 2008, ‘A computing system to assist business leaders in making ethical decisions’, in M. Oya, R. Uda and C. Yasunobu (eds.), Towards sustainable society on ubiquitous networks; Springer: Boston, 303-314.

Friedman, B., Kahn, P. H., Jr., & Borning, A.: 2008. ’Value Sensitive Design and information systems’, in K.E. Himma & H.T. Tavani (eds.), The Handbook of Information and Computer Ethics; John Wiley & Sons, Inc.: Hoboken, NJ, 69-101.

Gotterbarn, D. W.: 2002, ‘Reducing software failures: Addressing the ethical risks of the software development lifecycle’. Australian Journal of Information Systems 9(2), 155-165.

Kant, I. Grundläggning av sedernas metafysik; Daidalos: Stockholm, 2006.

Kohlberg. L.: 1985, ‘The just community: Approach to moral education in theory and practice’, in M. Berkowitz and F. Oser (eds.), Moral education: Theory and application; Lawrence Erlbaum Associates; Hillsdale, NJ.

Laaksoharju, M. and Kavathatzopoulos, I.: in press, ‘EthXpert: The basic structure and functionality of a decision support system in ethics. International Transactions in Operational Research.

Mancherjee, K. and A. Sodan,: 2004. ‘Can computer tools support ethical decision making?’. ACM SIGCAS Computers and Society, 34: 1.

Piaget, J. The moral judgement of the child; Routledge and Kegan Paul: London, 1932.

Szejko S.: 2002, ‘Incorporating ethics into the software process’, in I. Alvarez et. al (eds.), The transformation of organisations in the information age: Social and ethical implications, ETHICOMP 2002; Univeridade Lusiada: Lisbon, 271 – 279.

Comments are closed.