Incorporating Ethics into Computer Science Research Dr. Gary Comstock, Director, NCSU Research Ethics Program Dr. Thomas Honeycutt, Professor, Computer Science Dr. Ed Gehringer, Professor, Computer Science David Wright, NCSU Junior Research Ethics Fellow

Agenda † What is Research Ethics? † Four Approaches to Making Moral Decisions † Three Important Ethical Issues in CS Research † Case Study

05/01/2006

2

What is Research Ethics? † Ethics – the study of arguments about right & wrong, good & bad † Research – scholarly or scientific investigation or inquiry † Research Ethics – the study of arguments about right & wrong ways of conducting and reporting scholarly inquiry 05/01/2006

3

Research Ethics & Computer Science Computer are nearly ubiquitous Influence spreads even farther Research moves into application quickly Ripple effects are often impossible to predict † Demarcation of human subjects or participants in CS research often unclear † † † †

05/01/2006

4

Four Approaches to Making Moral Decisions † Consequentialist – results of a decision are what matters † Non-Consequentialist – outcome is secondary to following a general or accepted principle of action † Virtue – “What would a moral person do in this situation?” † Care Ethics – respect for thoughts, feelings, & relationships of all concerned is the highest good 05/01/2006

5

Three Important Ethical Issues † The Role of Software in Research † Rational Discourse, Replication, & Criticism † Human Participants in Computing Research

05/01/2006

6

The Role of Software in Research † Software as a Phenomenon for Study „ Implementation bias „ Programming quality „ Record keeping & documentation

† Metrics & Measurement with Software „ Choosing the right yardstick „ Measurement bias „ Tool interaction 05/01/2006

7

Rational Discourse, Replication, & Criticism † Demonstration vs. Empiricism

„ Is CS fundamentally different from other scientific disciplines? „ Is Empirical judgment & evaluation necessary?

† Duplication & Extension

„ How should research be documented? „ Are multidisciplinary perspectives necessary?

† Abstract Findings vs. Unstated Problems „ Majority of publications in CS & SE present abstract & formative findings „ Lack of rigorous experimental evaluation “unacceptable,” “alarming,” and an “embarrassment” 05/01/2006

8

Human Participants in Computing Research † Informed Consent „ Example: Open source software developers may not intend their work to be a subject of research

† Inadvertent Coercion „ Student “volunteers” in classroom research „ Recruitment by coworkers or supervisors in the workplace 05/01/2006

9

Research Ethics & Computer Science † Hall & Flynn, 2001, Survey investigating awareness of research ethics among academic CS & SE researchers in UK: „ “I find this questionnaire very worrying because the idea of having to seek ethical approval threatens academic freedom.“ „ “(Seeking ethical approval) has never arisen and I don't know why this is an issue." „ “No one is responsible for the ethical approval of CS research."

05/01/2006

10

A Case of Anomalies: Research Ethics & Computer Science † Can Ann & Ken ignore their anomalous results? † Should anomalous results be mentioned in publications? † Should prior comments regarding anomalies be accepted as is, or should research be extended to attempt explanation? † What steps could Ken & Ann take to further identify the suspected bias? † What risks do researchers expose themselves to by attempting (or not) to identify bias? † What risks exist for these students’ advisor if either course of action is taken? † What ethical rules of thumb offer guidance in this situation? 05/01/2006

11

Research Ethics @ NCSU † Research Ethics Homepage: http://www.chass.ncsu.edu/ethics/ † LANGURE Homepage: http://www.chass.ncsu.edu/langure/ † Motivation, Design, and Ubiquity: A Discussion of Research Ethics and Computer Science http://www.chass.ncsu.edu/langure/modules/documents/Draft6April2006.pdf

05/01/2006

12

Final Thoughts In the history of the development in technical change, very often the people responsible for a certain specialty are then followed by a technical innovation. And then the people who become responsible for the field after the technical innovation are a completely different group of people. Christopher Alexander, Address to OOPSLA ‘96 05/01/2006

13