EastonPsyOrf322S05Paper

advertisement
Marquis Martin-Easton
ORF/PSY 322
Final Project Paper
Can Computer Tools Support Ethical Decision
Making?
A Look at the Ethos Tool
With the rise of computer/information technology in today’s society there
has also been a rise in unethical decision making. Today the
computer/information technology such as the Internet has opened a floodgate of
immoral behavior. Plagiarism, illegal file sharing are just two examples of
unethical behavior that has increased due to advances in computer/information
technology. The rise in unethical decision making isn’t limited to the academic
and recreational platforms, in the business world there have been numerous
reports of unethical decisions being made: insider-trading convictions, the Enron
scandal, Intel’s refusal to recall defective CPU’s, etc. For businesses the
pressure to succeed and increase revenues leads to deliberately poor or
careless actions. In a society in which computer technology is so vital and where
ethical decision making is being questioned, those in the business world are
asking: Can computer tools and/or systems be used to support or oppose so
called ethical decisions?
Ethics in Computer Science
Along with the creation of the World Wide Web and the rise in plagiarism,
file sharing, viruses and overall abuse of computer/information technology, there
have also been the creation of computer societies and organizations such as the
ACM and the IEEE who have established codes of ethics for their members in an
effort to increase ethical awareness and support ethical decision making. But
what makes a decision ethical or unethical? In the case of computer science,
ethics and/or deciding what is ethical is the typical understanding that morality
includes two key components: value (personal or societal) and concern for
others. There are many different categories of ethics in computer science; from
using computers properly and protecting the rights of software designers to a
more business prospective of computer scientist involvement in contract
negotiations, hiring, and paper reviews. Computer scientist have been working
on creating computer tools and systems that implement “moral theories and test
procedures [to] provide a formalized framework which aid in identifying moral
reasons, relevant information, and available options for courses of actions.”1
Some these tools have, indeed, successfully implemented moral theories and
test procedures, leading computer scientists to be confident that computer tools
can identify moral dilemmas and, thus, assist in solving them.
Ethical Decision Making and Ethics Tools
There are many different ethical theories, however there are two main
schools of thought: consequence theories and respect-for-persons theories.
Consequence theories are mostly represented by utilitarianism (or utilitarian
theory), which evaluates consequences based on a measure of happiness
(utility). The goal of utilitarianism is to maximize the overall utility (happiness) for
the relevant addressees. On the other hand, respect-for-persons theories believe
1
Mancherjee, Kevin and Sodan Angela “Can Computer Tools Support Ethical Decision Making?”
that people are independent “moral agents” and that each of the relevant
addressees should be equally respected (including oneself). This type of respectfor-persons ethics is heavily based on Kant’s duty (what is right/responsible)
ethics, though some consider that it also includes Locke’s rights (human rights)
ethics. Obviously, there are other types of ethical theories (such as virtue ethics),
however the usefulness of utilitarianism and respect-for-persons ethics is that
they can be tested and applied to formalized procedures. Utilitarian theory has
the act utilitarian test, the cost/benefit analysis and the rule utilitarian test, while
respect-for-persons theory has the Golden Rule test, the self-defeating test and
the rights test.
The Utilitarian and the respect-for-persons theories have limitations. They
“either boil down all considerations to mainly monetary terms and potentially
violate the interests of the individual for the sake of the community, or potentially
[make] the individual an obstacle for the interests of the community.”2
Unfortunately, though, no one has been able to come up with a “master plan” to
eradicate the limitations in the theories, however, many computer scientist agree
that a combination of both theories is the best approach. There are ethical
decision aids that use either utilitarianism or respect-for-persons theory. SoDIS
(Software Development Impact Statements) employs a utilitarianism approach as
it collects and analyzes data associated with products or people. However, it
lacks the power to compare different social perspectives. The respect-forpersons ethics is utilized by the RESOLVDD and the Paramedic method.
RESOLVDD checks predefined moral values and compares them with each of
2
Mancherjee, Kevin and Sodan Angela “Can Computer Tools Support Ethical Decision Making?”
the parties involved in the problem, while the Paramedic method focuses an
audiences’ obligations. Unfortunately, both RESOLVDD and the Paramedic
method don’t assess the value of the possible decision outcomes. The Ethos
Tool, on the other hand, applies the utilitarian theory with the respect-for-persons
theory for a better, more thorough, assessment of possible courses of action.
The Ethos Tool
The Ethos Tool combines and implements both the utilitarian and the
respect-for-persons theories while applying the HARPS Ethical Analysis
Methodology. The Ethos tool utilizes “a guideline approach, designed to aid in
the consistent application of ethical codes by creating a framework with clear
procedural steps to address ethical problems in a rational manner and make the
application of one’s moral values and beliefs consistent.”3 The Ethos tools’ user
interface is presented as a map and is always visible to the user while the
program is operating. The users past answers are stored within in the tool, which
creates a history of the users decisions allowing for others to review and judge
the rationale of the operator’s past decisions. The tool also allows users to revisit
and add more information that will assist the tool in refining the problem and
providing better possible outcomes.
The HARPS Methodology used by the Ethos tool has three basic steps:
frame the problem, outline the alternatives and evaluate the alternatives. Once
the tool initiates all three steps in the HARPS Methodology it produces an
argument for the course of action it deems the most ethical. The first step is:
3
Mancherjee, Kevin and Sodan Angela “Can Computer Tools Support Ethical Decision Making?”
Framing the problem. In this step, the user defines the problem for the Ethos tool.
The tool examines the problem and the user inputs “the role of the moral agent
(the user), known and unclear facts (facts and factual issues), potentially unclear
definitions (conceptual issues) and the open moral questions (moral issues).”4
The Ethos tool then moves on to the second step: Outline the alternatives. After
posing the problem to the tool, the user must now develop possible solutions.
Information on all requirements (duties toward certain individuals, responsibilities
as an employee, etc.) towards the parties involved in the problem is collected.
The tool separates ethical issue problems into two categories: line drawing
(defining what is/isn’t acceptable behavior) or conflict resolution (two or more
requirements are in conflict). Once the issues are separated, the Ethos tool
begins to assign weights to the different arguments and generates possible
solutions/actions that warrant further examination in the next step of the HARPS
Methodology: Evaluate the alternatives. Now that the tool has a group of
solutions to work with, it begins evaluating each one with a series of utilitarian
and respect-for-persons analyses. The utilitarian analyses includes: act utilitarian
analysis, which gives ratings to the impact each action has on the different
parties involved; cost/benefit analysis (used when there’s a plethora of factual
and statistical information), which gives each solution a measurable value
(usually it’s a monetary value); and rule utilitarian analysis which “checks if the
different acts would be suitable if generalized as a rule for society, establishing
4
Mancherjee, Kevin and Sodan Angela “Can Computer Tools Support Ethical Decision Making?”
both the rule and its alternative, and thinking of the consequences.”5 The
respect-for-persons analysis includes: rights-for-persons check, which ensures
that the rights of each individual involved are not being violated by any of the
possible actions; golden rule test, which confirms that each individual is being
treated fairly and equally in each of the actions; and self-defeating test, which
guarantees that each the possible actions are not self-contradicting. Because the
tests are dealing with two different ethical theories, the Ethos tool is bound to
come across conflicting ethical obligations. In the case of the occurrence of
conflicting ethical obligations generated by the Ethos tool, the tool will rank each
conflict against each other based on usefulness.
Once the Ethos tool has made its way through each step of the HARPS
Methodology, it gives the solution(s) “that satisfies the greatest extent of the
obligations/requirements by overall greatest usefulness.”6 Accompanying the
solutions is an explanation of the rationale behind the tools choosing of a
particular course(s) of action(s).
Does the Ethos Tool Truly Aid in Making of Ethical Decisions?
The Ethos Tool has been tested in a case study involving faked numbers
in contract bids. The tool performed well, in the sense that it correctly evaluated
the problem and produced an ethically sound solution. However, participants in
the study were reluctant to use the tool because it was very time consuming,
there were too many manual inputs; and at times the tool confused the user,
5
6
Mancherjee, Kevin and Sodan Angela “Can Computer Tools Support Ethical Decision Making?”
Mancherjee, Kevin and Sodan Angela “Can Computer Tools Support Ethical Decision Making?”
participants had trouble understanding what the tool was asking them to do. One
of the main problems with the Ethos tool is the fact “that the moral tests
incorporated in the HARPS Methodology are only as effective as the moral basis
of the user…This means, the framework that the moral tests provide is as weak
or strong as the person using it.”7 Which is not a particularly good thing, if one is
attempting to make an ethical decision tool to help those who are not ethically
sound. Because of its time constraints and confusing interface, the Ethos tool
would be accepted more in a high school or university setting. In an academic
class, time isn’t a problem and ethical terminology and theories can be explained
before the actual use of the tool. To combat the fact that the Ethos tool does not
further the ethical standards of its user, there needs to be work on the tools
interface and should maybe be installed with a predetermined set of moral and
immoral actions/problems/etc. so that the possible moral corruptness of a user
has less of an effect on the ethical solutions produced by the tool. In its current
state the Ethos tool does indeed provide acceptable ethical solutions to morally
ambiguous problems, however it only provides ethical solutions to those who
already have a strong ethical foundation. Major improvements need to be made
to the Ethos tool’s interface and usability so it can truly be regarded as a viable
option in ethical decision making.
7
Mancherjee, Kevin and Sodan Angela “Can Computer Tools Support Ethical Decision Making?”
Download