Vidic 2:00 L18 COMPUTER TUTORS AND ETHICS Anna Siryk (ans215@pitt.edu) INTRODUCTION: AN ETHICALLYCHARGED SCENARIO IN DEVELOPMENT OF A COMPUTER TUTOR Suppose that I am a software engineer that is part of a team researching the effectiveness of spoken dialogue computer tutors. In the development of a computer tutor, I program a section that allows the software to track whether the student is using the Internet to solve questions, but only when the student is using the computer tutor at the same time. In later testing of the software, I have to decide whether or not information about that addition should be revealed to the students working with the computer tutor. In the interest of testing the software, it would be beneficial for me to not reveal this addition to the students, because it looks for behavior that changes the way the computer tutor reacts to a student’s answer. It has two important functions that are vital to the success of the computer tutor: testing for uncertainty and testing for boredom. Kate Forbes-Riley and Diane Litman of the Learning Research and Development Center at the University of Pittsburgh tested users’ performance using a computer tutor with and without uncertainty detection. They concluded that when uncertainty is recognized and adapted to, the local performance of the user improves significantly [1]. Therefore, in the context of this scenario, if a student is unsure of their answer to a question, and they find the answer using the Internet, their answer will sound certain, while they are actually unsure. Consequently, the computer tutor will not be able to respond accordingly, which will worsen the student’s performance. Tracking students’ usage of the Internet while using this tutoring software can also be used to infer whether they are bored or frustrated. Boredom has been shown to correlate with less use of cognitive strategies while learning, and frustration may disengage the student from a task and reduce learning overall [2]. Hence, it is essential to check for signs of these states so the software can adjust its output to prevent these negative outcomes. In light of this information, would it be ethical if I chose to not inform the students of the software’s ability to track their Internet usage, if I wanted to ensure the best results from testing? I am going to use the National Society of Professional Engineers’ (NSPE) Code of Ethics, the Software Engineering Code of Ethics, and articles from accredited professors of ethics and professors of engineering to answer this question. However, in this particular scenario, are all of these resources helpful in making an ethical decision? University of Pittsburgh, Swanson School of Engineering 2013-10-29 APPLYING THE NSPE CODE OF ETHICS The first section of the NSPE Code of Ethics that may apply to this situation is Section I.5, which simply states, “Engineers, in the fulfillment of their professional duties, shall avoid deceptive acts [3].” Withholding information can be seen as a deceptive act, therefore, if I just consider the action of omitting information from students, I could say my behavior is unethical. However, since the reasoning behind this action is to test the software’s effectiveness, I do not think this section pertains to this situation. If I did not withhold this information, the research would likely be skewed, because the students would most likely be biased against using the Internet if they knew that the computer tutor would recognize their attempts at cheating the system. Section III.3.a of the NSPE Code of Ethics, which is the only other section that may also apply to this situation, states that “Engineers shall avoid the use of statements containing a material misrepresentation of fact or omitting a material fact [4].” If I, as a software engineer, chose not to inform students of an invasion of their privacy, I would not be adhering to this statement. However, since I would be omitting a fact while testing a product rather than while selling one, omitting this information would not detriment the students in any way, since they would be using this program while under observation, which would not allow them privacy anyhow. Therefore, I do not believe this statement is particularly applicable to this scenario. Unfortunately, the NSPE Code of Ethics is too broad to cover this particular scenario. Adding in the factor of being part of a research team makes the statements in this code of ethics too ambiguous. As a consequence, they cannot be used to accurately decide whether or not withholding information from students that test this software is unethical. APPLYING THE SOFTWARE ENGINEERING CODE OF ETHICS The decision to not inform students testing a program that a computer tutor knows when they access the Internet deals primarily with software engineering. Therefore, the Software Engineering Code of Ethics will contain more specialized guidelines pertaining to this situation. Additionally, this code of ethics has more statements that will apply specifically to me, as a software engineer. Principle 1.06 states, “Be fair and avoid deception in all statements, particularly public ones, concerning software or related documents, methods and tools [5].” This statement is generally the same as the NSPE Code of Ethics Section I.5, in that an engineer should “avoid deception.” Therefore, like the Anna Siryk NSPE Code of Ethics, this statement does not clarify whether withholding certain information from students that test a program would be considered “deception.” Again, this statement is too broad to cover this particular scenario. Principle 3.01 of the Software Engineering Code of Ethics states that software engineers should “strive for high quality, acceptable cost and a reasonable schedule [6].” Informing students of the software’s capability to check for Internet use would bias them against their natural tendencies, and would skew the results. By this principle, the decision to keep the testing of the software unbiased would be ethical, because it would ensure the highest quality product. This code of ethics goes on to state, in Principle 3.10, that software engineers should “ensure adequate testing, debugging, and review of software and related documents on which they work [7].” As stated above, “adequate testing” of the computer tutor would not involve informing students of all of its components, as that would bias them, resulting in the testing not accurately showing how well the program works. Therefore, by this principle, my decision is ethical. Principle 3.12, however, states that software engineers should “work to develop software and related documents that respect the privacy of those who will be affected by that software [8].” In this scenario, the computer tutor that I am programming will check for specific Internet use while the program is running, which could be seen as not respecting the privacy of the students using the software. In that sense, my decision to not tell students testing the program about this capability would be unethical, since the software would not be respecting their privacy. As expected from its more specialized statements, the Software Engineering Code of Ethics provides much more helpful ethical values than the NSPE Code of Ethics. From the principles of it that pertain to this scenario, I conclude that my decision is ethical. While the computer tutor may slightly invade the privacy of the students, which violates Principle 3.12, accurate testing, which is necessary to uphold Principle 3.01 and 3.10, holds a higher ethical priority. Therefore, the need for unbiased testing of the software overrules any slight invasion of privacy that will occur. from the students agreeing to test my software, it could be argued to be a case of “nondiligence,” since the students would be making an agreement without knowing all the facts. To avoid such unethical behavior, however, I could simply include a statement in the agreement stating that the students could not be informed of every aspect of the computer tutor software for testing purposes, and let them act as they would. Peter Froelich, a lecturer in the Department of Computer Science at Johns Hopkins University, argues that when engineers rely entirely on codes of ethics, it rarely provides a completely ethical solution, because there are so many details that should be common sense, but are not considered due to not being listed in these codes of ethics [10]. I believe this is a fair point. So, by looking at my decision from my own perspective, I believe that not informing the student testers of my software about the computer tutor’s ability to check for Internet use is completely ethical. In the same way a teacher should know if their students are using the Internet during a test, I believe the computer should know if a student is using the Internet during their sessions. Kenneth D. Pimple, Director of Teaching Research Ethics Programs at the Poynter Center for the Study of Ethics and American Institutions, holds the view that computing systems’ small invasions of privacy are not, by themselves, unethical [11]. Therefore, my computer tutor’s brief Internet usage checks should not be unethical, since they are just small, limited invasions of privacy. In a more general sense of software engineering ethics, Deborah G. Johnson, a professor of applied ethics at the University of Virginia, stated that the basis for ethical decisions should be “trust.” Specifically, decision makers should act in a way so as to be worthy of public trust [12]. Applying this logic to my decision, I believe that if I made the decision to withhold a piece of information in order to have more accurate results in testing, I would gain more trust than if I gave away the information and skewed the results. CONCLUSION: ETHICAL DECISION MAKING: THIS AND OTHER SCENARIOS After consulting a multitude of sources to decide whether or not I should inform the student testers of my software’s ability to track Internet usage, the ethical decision would be not to inform them. Although tracking students’ Internet usage, even for a short period of time, is unethical under Principle 3.12 of the Software Engineering Code of Ethics, which states that engineers should strive to develop software that does not invade the privacy of its users [4], the code of ethics goes on to state that accurate testing and high quality products are also a priority. The issue of “nondiligence,” which is an agreement made without full understanding of what is being agreed upon [9], can be rectified with simply stating that the students testing the software cannot be told the full details due to testing purposes. As I also believe that my OTHER ETHICAL CONSIDERATIONS Since the NSPE and Software Engineering Codes of Ethics are general, other sources of ethical guidelines should be considered for this situation. One such source is the article that Brian Berenbach of Siemens Corporate Research and Manfred Broy of the Technical University of Munich published, which elaborates on more specific ethical dilemmas in software engineering than the Software Engineering Code of Ethics provides. One of these dilemmas, which the authors dubbed “nondiligence,” states that a situation in which agreements are made without a clear understanding of what is being agreed upon is unethical [9]. Therefore, if I decided to withhold information 2 Anna Siryk decision is morally correct, I have no reservations about judging this decision to be ethical. I advise engineers who are looking to act with ethical integrity to first consult their own moral compass, because most codes of ethics are based on “common sense” of what is right. Then, I recommend first consulting the code of ethics of the specific type of engineering the decision deals with, and then referring to the NSPE Code of Ethics, because those specialized codes of ethics contain more specific guidelines, which are more likely than the NSPE’s to contain relevant information. Additionally, some of the most thought-out ethics guidelines are those published by professors of ethics and professional engineers, and so specific problems one might encounter may already have been thoroughly debated in engineering publications. [8] IEEE-CS/ACM Joint Task Force on Software Engineering Ethics and Professional Practices. (1999). 3.12. Software Engineering Code of Ethics and Professional Practice. http://www.computer.org/portal/web/getcertified/resources/c ode_of_ethics [9] B. Berenbach, M. Broy. (2009). “Professional and Ethical Dilemmas in Software Engineering.” Computer. (Online Article). http://ieeexplore.ieee.org.pitt.idm.oclc.org/stamp/stamp.jsp?t p=&arnumber=4755159&tag=1 [10] P. Froelich. (2008). “Ethics considered harmful.”ACM SIGCAS Computers and Society. (Online Article). http://delivery.acm.org.pitt.idm.oclc.org/10.1145/1390000/1 386587/p21froelich.pdf?ip=132.174.255.116&id=1386587&acc=ACTI VE%20SERVICE&key=C2716FEBFA981EF17A2771D9E 7C32E8FAD577D985794B35D&CFID=365222182&CFTO KEN=78882433&__acm__=1382832089_aa8eca8451537a6 d27d23ab5ad148be4 [11] K.D. Pimple. (2011). “Computing Ethics Surrounded by Machines.” Communications of the ACM. (Online Article). DOI: 10.1145/1897852.1897864. [12] D.G. Johnson. (2008). “Computing ethics: Computer experts: guns-for-hire or professionals?” Communications of the ACM. (Online Article). DOI: 10.1145/1400181.1400190 REFERENCES [1] K. Forbes-Riley, D. Litman. (2011). “Benefits and challenges of real-time uncertainty detection and adaptation in a spoken dialogue computer tutor.” Speech Communication. (Online Article). http://www.sciencedirect.com.pitt.idm.oclc.org/science/articl e/pii/S0167639311000318?np=y [2] M.M.T. Rodrigo, R.S.J.D. Baker, J. Agapito, J. Nabos, M.C. Repalam, S.S. Reyes Jr., M.O.C.Z. San Pedro. (2012). “The Effects of an Interactive Software Agent on Student Affective Dynamics while Using an Intelligent Tutoring System.” IEEE Transactions on Affective Computing. (Online Article). http://ieeexplore.ieee.org.pitt.idm.oclc.org/stamp/stamp.jsp?t p=&arnumber=6122013&tag=1 [3] National Society of Professional Engineers. (2007). I.5. NSPE Code of Ethics for Engineers. http://www.nspe.org/Ethics/CodeofEthics/index.html [4] National Society of Professional Engineers. (2007). III.3.a. NSPE Code of Ethics for Engineers. http://www.nspe.org/Ethics/CodeofEthics/index.html [5] IEEE-CS/ACM Joint Task Force on Software Engineering Ethics and Professional Practices. (1999). 1.06. Software Engineering Code of Ethics and Professional Practice. http://www.computer.org/portal/web/getcertified/resources/c ode_of_ethics [6] IEEE-CS/ACM Joint Task Force on Software Engineering Ethics and Professional Practices. (1999). 3.01. Software Engineering Code of Ethics and Professional Practice. http://www.computer.org/portal/web/getcertified/resources/c ode_of_ethics [7] IEEE-CS/ACM Joint Task Force on Software Engineering Ethics and Professional Practices. (1999). 3.10. Software Engineering Code of Ethics and Professional Practice. http://www.computer.org/portal/web/getcertified/resources/c ode_of_ethics ADDITIONAL SOURCES S. Valentine, F. Vides, G. Lucchese, D. Turner, H. Kim, W. Li. (2013). “Mechanix: a sketch-based tutoring and grading system for free-body diagrams.” AI Magazine. (Online Article). http://go.galegroup.com.pitt.idm.oclc.org/ps/i.do?action=inte rpret&id=GALE%7CA325892061&v=2.1&u=upitt_main&i t=r&p=AONE&sw=w&authCount=1 B.M. McLaren, K.E. DeLeeuw, R.E. Mayer. (2011). “A politeness effect in learning with web-based intelligent tutors.” International Journal of Human-Computer Studies. (Online Article). http://www.sciencedirect.com.pitt.idm.oclc.org/science/articl e/pii/S1071581910001205?np=y P.A. Jaques, H. Seffrin, G. Rubi, F. de Morais, C. Ghilardi, I.I. Bittencourt, S. Isotani. (2013). “Rule-based expert systems to support step-by-step guidance in algebraic problem solving: The case of the tutor PAT2Math.” Expert Systems with Applications. (Online Article). http://www.sciencedirect.com/science/article/pii/S09574174 13002418 A. Leuski, D. Traum. (2011). “NPCEditor: creating virtual human dialogue using information retrieval techniques.” AI Magazine. (Online Article). http://go.galegroup.com.pitt.idm.oclc.org/ps/i.do?id=GALE %7CA261385406&v=2.1&markList=true&u=upitt_main&it =r&p=AONE&sw=w 3 Anna Siryk ACKNOWLEDGEMENTS I would like to wholeheartedly thank my writing instructor, Liberty Ferda, for ruthlessly reviewing my work to improve my writing, and I would also like to thank my brother, Alex Siryk, for ensuring that I persevere, and for supporting me in his own way. 4