IN PURSUIT OF ETHICAL IS RESEARCH
Gove Allen, Nick Ball, Jeff Smith
Working Paper, please do not cite without authors’ permission.
ABSTRACT
IS researchers frequently face ethical quandaries in their professional lives. Appeals to existing codes of professional conduct, such as the Association for Information Systems
(AIS) code of professional conduct often offer little in the way of practical guidance. We present the results of a study of IS researchers that assesses their attitudes and the prevalence of 29 ethically questionable behaviors. We find that career focus, culture, AIS membership, and the existence of the AIS code of professional ethics are associated with
IS researchers’ perceptions of these behaviors. We further find that respondents report that they are more likely to engage in these behaviors in the future than they report to have engaged in them in the past. As a result of the study, we recommend that the IS community revisit the AIS code of professional ethics, take active steps to educate its members on professional ethics, and take steps to actively uphold the ethical standards of our community.
INTRODUCTION
In July 1999, Ned Kock (1999) published “A Case of Academic Plagiarism” in the Communications of the ACM , and this served as a “call to arms” for many in the information systems (IS) academic community. By recounting his traumatic experiences as a victim of plagiarism, Kock made us aware of an area of extreme vulnerability – and one in which, it became apparent, we were not of one mind regarding the rules.
By late 2003, consistent with recommendations in a follow-up “Issues and
Opinions” article in MIS Quarterly (Kock and Davison, 2003) the Association for
Information Systems (AIS) released its “AIS Code of Research Conduct,” which addressed issues of plagiarism and a few other research issues such as rights of research subjects. To be sure, other related codes (e.g., Academy of Management, Association for
1

Computing Machinery) were already available at the time, but the development and dissemination of the AIS code highlighted many issues directly for the IS academic community.
In the ensuing years, our experiences and our discussions with IS authors, reviewers, and editors have led us to question whether our academic community has come to agreement on research-related behaviors. And, even in some areas in which it appears that we have general agreement, observers frequently cite examples in which violations are alleged to have occurred.
In order to address our own curiosity about possible gaps in both agreement and implementation, we undertook a survey exercise in 2008. With endorsements from the president of the AIS and from the editors-in-chief of Information Systems Research , the
Journal of AIS , and MIS Quarterly , we developed and distributed an online survey to IS researchers around the world.
In the sections below, we first describe briefly the theoretical research domain into which this survey exercise falls. We then discuss a set of findings based on the survey responses, we consider some challenges associated with those findings, and we provide our own suggestions for addressing some phenomena that have been uncovered.
PROFESSIONAL DUTIES
Human beings face quandaries in many areas of life: we must often decide “what is the right thing to do?” These quandaries occur in many domains including our personal lives, our workplaces, or our relationships with society and the world. We
2

frequently search for normative guidance, and we can appeal to several bodies of theory for help.
Although traditional philosophical ethics (see Frankena, 1973 and Laudon, 1995) and the theories of business ethics (see Smith and Hasnas, 1999) provide normative guidance for some activities, IS researchers are more likely to be enlightened in answering “what is the right thing to do?” by considering their professional duties . This is the case because many actions that are clearly right or wrong from a professional perspective are not as visibly illuminated in light of traditional philosophical ethics. To the extent that answers can be found, they will most likely reside in sets of stated or unstated rules that have been established or inferred at a professional level. In some professions (e.g. law and medicine) these rules are very well articulated, in others (e.g. IS academic research) they are less so. Sometimes called “professional codes of ethics” or
“codes of conduct,” these sets of rules differ somewhat from normative theoretical approaches because they are usually quite specific and are intended to reduce the amount of interpretation at the individual level. They also serve to inform individuals on actions that may not otherwise be clear from a normative perspective. In one sense, referring to them as rules of “ethics” is somewhat misleading, since the word is used in a very different context than in philosophical theory. The “codes” of professional duties are best viewed as agreed-upon “rules of the road” for behavior in a certain profession. They often overlap with the normative obligations that are deduced from the philosophical theories but this is not guaranteed; they may even be in conflict with the other theories in their assertions regarding the “right” actions.
3

The broad profession of “academic researchers,” and the somewhat narrower profession of “IS academic researchers” both embrace documented “codes” that define some of these normative rules. IS researchers might appeal to the code of the Academy of Management (AoM) or the code of the AIS (Academy of Management Revised code of Ethics, 2008; Association for Information Systems Code of Research Conduct, 2008).
As in many professions, there may also be unstated rules or assumptions for IS academic researchers. These might be institutionalized in other ways, but could be as binding on individual behaviors as the codified rules of professional associations.
While the domain of professional duties can provide explicit direction for those in the profession, there are some obvious challenges associated with this approach to influencing behavior. One such challenge is associated with the need for those within the profession to acknowledge the legitimacy of the rules. In some professions (e.g., law), this is enforced through required membership in an association. In such cases there are meaningful penalties for violating the rules. In others (e.g., academic research), there is no such membership requirement, so it is more difficult to ensure that all members of the profession acknowledge–or even be aware of–a particular set of rules. A second challenge associated with the domain of professional duties is that, even though a profession may attempt to codify its expectations at a level far more specific than the philosophical theories, it remains challenging to produce an exhaustive set of rules that will address all professional situations.
It is in this vein that we studied IS academic researchers’ perspectives on appropriate behaviors within the research domain. We focused on the extent to which members of the IS research profession have agreed upon the normative principles–either
4

codified or based on unstated rules or assumptions–that are expected to guide professional behaviors. The details of our study’s methodology are highlighted in
Appendix 1; the individual behaviors we studied can be found in Appendix 2.
We now discuss the important findings from our survey.
SURVEY FINDINGS
As part of the study, we examined 29 research behaviors in five areas: (1) authorship and authorship order, (2) conducting research, (3) reporting on research, (4) serving as a reviewer, and (5) handling papers as an editor. For each of the 29 behaviors, respondents were presented a short hypothetical scenario that depicted one of the research behaviors and asked to make assessments of the appropriateness and pervasiveness of the behaviors described in the scenario. The mean scores and standard deviations for these assessments can be found in Appendix 3.
Respondents were also asked to provide some background and demographic information. This allowed us to examine the set of responses to determine what elements of the respondents’ experience may influence their attitudes and perceptions of the behaviors. From this analysis there are five major findings: (1) professional emphasis matters, (2) future behavior is projected to be less ethical that past behavior, (3) culture matters, (4) AIS membership matters, and (5) the AIS Code of Research Conduct matters.
Professional emphasis matters
To assess the extent to which professional emphasis influences our attitude and perceptions of research ethics, we asked the respondents to assess the percentage of their time they spent in research related activities, teaching related activities, and service related activities. In segmenting the responses we divided the sample based on the
5

percentage of time that the respondents spent conducting research. Those that reported spending 50% or more of their time on research were placed into our high-research group. Those that reported spending 40% or less of their time on research were placed in our low-research group.
We then conducted a series of t-tests to examine differences in the mean inappropriateness scores 1 for each of the 29 research behaviors between the high-research and low-research groups. Table 1 outlines the results of this analysis. We found significant differences in the mean inappropriate scores for 12 of the 29 behaviors. In each case where significant differences were found, the low-research group demonstrated a higher inappropriateness mean. In other words, when there were differences in opinion between the groups on how inappropriate a behavior was, the group spending less of their time on research reported that the behaviors were more egregious than the group that was more engaged in research.
It appears that when individuals becomes more heavily engaged in the research process, they become pragmatic about research ethics-related issues. Given a strong relationship between attitude and behavior, this finding is concerning. We worry that that those members of our community that are most engaged in research are also less concerned about the appropriateness of potentially unethical research practices. This may be even more alarming when considered in juxtaposition with our next key finding.
If you don’t like what we are currently doing, just wait…
We examined the correlation between our normative assessments of the research behaviors and reported past engagement in those behaviors and found a significant
1
Respondents were asked to evaluate how “appropriate” described behaviors were on a 1-7 scale where 1 is “clearly appropriate” and 7 is “very inappropriate”. Because the higher the response value, the more inappropriate the assessment, we term this measure “inappropriateness” score.
6

positive relationship. Our respondents tend to engage more in those practices they deem to be more appropriate. This is not a particularly surprising or compelling finding. We then examined the relationship between respondents’ reported past engagement in the behaviors and the extent to which they expect to engage in similar behaviors in the future.
We conducted a set of t-tests to determine if there were differences in the mean scores between the reported past engagement in the behaviors and the intended future engagement in the behaviors. We found that for all 29 behaviors the mean score for the intended future behavior was higher than the mean score for reported past behavior. This is a striking and counter-intuitive result. Regardless of our respondents’ assessment of the appropriateness of the behaviors, they report that they are more likely to engage in each behavior in the future than they report engaging in the behavior in the past.
Although we can only speculate on the root of this finding, it does raise an important issue for our field. If our respondents’ reported intentions correlate with actual future behaviors, individuals in our field are likely to engage in each of these potentially unethical practices more in the future than they currently do. It could be that as a filed, we will become less ethical over time.
Culture matters
Beyond professional emphasison research and our future intentions, we examined the extent to which culture influences our respondents’ attitudes concerning the appropriateness of the research behaviors. To capture a person’s cultural background, asked respondents to report where they received their childhood education. Respondents were then segmented into two group: those who were educated in North America, and those who received their childhood education outside of North America. Our results
7

suggest that culture does play a role in influencing our perceptions of the appropriateness of the research behaviors.
Again, t-tests were conducted to determine differences in the mean scores between those educated in North America and those educated outside of North America.
We found significant differences for 12 of the 29 behaviors. Within these 12 differences, an interesting pattern emerged. Three of the 12 behaviors related to authorship issues.
Although there was agreement on the egregiousness of dropping a coauthor from a paper, those educated outside of North America report that adding a coauthor or manipulating authorship order for reasons other than contribution to the work are more inappropriate than those educated in North America. With one exception, the remaining nine behaviors were all in the other direction. Those educated in North America report that the behaviors are more inappropriate.
We can only speculate on the reasons for this finding. Whatever the cause, it appears that our culture does in fact influence our perceptions of ethics. This finding has important implications for our field. There are aspects of our past that influence our perceptions–any effort to arrive at consensus in the field on research ethics must consider individuals’ varied cultural backgrounds.
AIS membership matters
In addition to culture, it appears that our AIS community may have an influence on our perceptions as well. As part of the questionnaire, we asked respondents to report whether they had ever been a member of the AIS. We segmented our data on AIS membership to determine if there were significant differences between those who had been a member of the AIS and those who had not been.
8

T-tests were performed to discover if significant differences were present between the two samples. In total, the mean appropriateness scores for 13 of the 29 behaviors are significantly different. Of these differences, in most cases (9/13) those who had been AIS members reported higher levels of inappropriateness for the behaviors described.
Strikingly, as with our analysis of culture, three of the behaviors that non-AIS members found to be more egregious dealt with authorship issues. This is despite the fact that we find only a relatively modest relationship (coefficient of correlation of 0.29) between
AIS membership and region of childhood education.
Although it may be difficult to determine what drives the association between attitudes about ethics and AIS membership, it is clear that there is some relationship. In fact, this association is generally positive for our community. In our efforts as a field to promote ethical behavior, one important avenue for success may well lie in leveraging the influence of the AIS.
The AIS Code matters
In addition to examining how the individual characteristics of respondents influence their attitudes and intentions about the behaviors, we also examined the relationship between the AIS Code of Research Conduct and our responses to the questionnaire. Table 2 highlights this analysis. Our findings suggest that there is a relationship between the attitudes and intentions of our respondents and whether individual behaviors are institutionalized in our formal AIS code.
To examine this question, we mapped our list of scenarios to the AIS code of conduct. Each behavior was scored on a three-point scale (0 to 2) based on how explicitly the AIS code address that issue. A score of 0 was assigned all behaviors where the AIS
9

code of research conduct was completely silent. A score of 2 was assigned to behaviors that are explicitly addressed by the AIS code. A score of 1 was assigned all behaviors where the AIS code addressed an issue vaguely or addressed a related behavior. We then examined differences in the mean scores for each set of behaviors.
In terms of inappropriateness, those items where the AIS code is silent have a significantly lower mean score (are viewed as most appropriate) than the behaviors coded
1 or 2. On the other hand, when the AIS code is explicit (items coded a 2), the mean inappropriateness score is significantly higher than the other items. These items are viewed as the most inappropriate items in the group.
Next we examined respondents’ reports of past engagement in the behaviors. Not surprisingly, respondents report a statistically higher level of past engagement in those behaviors that are not handled (scored a 0) in the AIS code. Interestingly, respondents also report a significantly higher level of past engagement for the items that are explicitly addressed (scored a 2) in the AIS code than for those items scored a 1. This may suggest that some behaviors are specifically addressed in the AIS code because members of the community have engaged in these behaviors in the past.
Finally, we examined the relationship between the AIS code and our respondents’ reported future intentions for engaging in the behaviors. We find that the mean scores for those items where the AIS code is silent are significantly higher than for the other behaviors. Further, we did not find a significant difference for future intentions on those items where the AIS code is vague or explicit. Our respondents report a lower intention to engage in these behaviors in the future for those items that are at least marginally addressed in the AIS code that for those items on which the code is silent.
10

In summary, the AIS code does appear to have an association with our responses.
Whether the AIS code is shaping or reflecting our opinions and intensions is unclear, but the relationship is incontrovertible. Although we are hesitant to suggest that the AIS should codify and regulate all possibly unethical research behaviors, it seems clear that the AIS Code of Research Conduct has a place in promoting ethical practices in our field.
RECOMMENDATIONS
Survey respondents report that they are more likely in the future to engage in some behaviors that are apparently viewed as questionable; this suggests that there should be a sense of urgency associated with our current situation.
In attempting to react to this sense of urgency, however, we may face some frustration because external factors may lead to non-convergence in our normative assessments. If the objective is to maximize consensus regarding our perceptions of appropriate and inappropriate behaviors, it may be quite difficult to reach agreement on which behaviors should, and should not, be codified into our professional rules.
Even so, we believe that our IS research community must confront the numerous areas in which we are either behaving at odds with our stated codes, in which we disagree with our codes, or in which our codes are silent. In that light, we offer the following recommendations to our research community.
We view the appropriate action plan as including three steps that are largely sequential but also overlapping: (1) revisit our AIS code, (2) educate the IS academic research community, and (3) develop mechanisms to uphold our community’s ethical standards. We discuss each.
11

Revisit our AIS code
Although there are several codes to which IS researchers might appeal as they seek to determine the “right” actions, the most obvious reference is the “AIS Code of
Research Conduct.” In light of the finding that there is a correlation between the explicitness of the code and member’s attitudes, behaviors and intentions, we recommend using the AIS code as a vehicle to influence our community’s behavior. However, determining the extent of what belongs in the code is not a simple matter. Clearly, the results of this study indicate that our field lacks agreement about the appropriateness of many research behaviors. Yet even if our field exhibited substantial agreement, this would not necessarily dictate what the code should include. It seems that the role of an association’s code of conduct is not to reflect the attitudes of the association’s membership; rather, the code should inspire ethical behavior. But given such disagreement among the membership, how does an association determine the content of its code of conduct?
The obvious mechanism for such an examination is a task force of IS academic researchers. Given our findings from our survey exercise, it is obvious that the task force membership should be comprised of researchers who span the spectra in terms of their cultural background and their research experience. Differences in assumptions and expectations exist across both of these dimensions, and the task force must have input from all perspectives.
We further recommend that the revised Code address the breadth of research, review, and editorial activities highlighted in this study (see Appendix 2). Although IS researchers agree on the inappropriateness of many behaviors, such as classical
12

plagiarism, there is less agreement regarding many others. A major objective of the task force effort should be to provide clear guidance in all the areas in which IS researchers encounter quandaries on a regular basis.
One important question is “to what extent must consensus be reached?” If the perspectives of task force members mirror those of the IS research community at large, it is likely that there will be disputes in some areas regarding the appropriateness of certain behaviors–with disagreements frequently based on career focus and cultural background.
While one might specify an algorithm such as “defer to the wisdom of those who have been doing this the longest” (apparently the AIS assumption; see AIS, 2008, “Good
Advice” section) or “defer to the cultural region in which the majority of IS researchers reside” (presumably, North America), such approaches strike us as overly simplistic.
We argue instead that a logically consistent set of rules can be best deduced by, first, agreeing on basic principles and, second, drilling down from those principles into specifics. Although it is beyond the scope of our present effort to complete this task, an example may prove helpful: Suppose that there could be general agreement on the principle that “IS researchers should respect the rights of human subjects.” This principle would then lead to rules regarding the necessity for complying with all local IRB registration and reporting requirements. As we saw in the survey results, many members of our IS research community do not at present comply with such requirements in all situations. The process of drilling down from the principle to the rules should – if documented and distributed – provide the necessary explanatory linkage to defend the codified rules for those in our community who are skeptical or cynical.
Educate the IS academic research community
13

Assuming that agreement has been reached on the Code, it will behoove our community to provide practical educational opportunities for three different audiences:
(1) doctoral students, (2) junior faculty, and (3) more senior researchers.
Although it is true that “wisdom can’t be told” (Gragg, 1940), it is nevertheless clear that the first step in developing wisdom regarding research behaviors is to engage in a process of enlightenment and acculturation during our doctoral training. We recommend that the AIS sponsor the development of a series of case-based materials that can be used as discussion vehicles in doctoral seminars. In fact, based on some responses we have received, the scenarios that were used in this survey may provide some fodder for initial efforts in this area. We envision a set of instructor materials being included in the package, with the expectation being that the various cases would be used to illustrate areas in which established rules clearly proscribe behaviors – and the ones in which researchers must interpret the rules or make their own judgments based on fundamental principles.
If all our doctoral programs eventually incorporate such training, then we can expect junior faculty members to be appropriately acculturated when hired. However, this is unlikely to occur for several years. Thus, in the meantime, we recommend a concerted effort to provide additional training for junior IS faculty. The “MIS Camps”
(held each year before some of the major IS conferences) would be an ideal venue for sessions that use a subset of the materials that are developed for the doctoral seminars.
It is also true that more senior research colleagues may benefit from a “refresher” course, especially in light of what will likely be additional codification of rules in the
14

near future. One way to accomplish this would be to have panel discussions in the major conferences where the proposed code of ethics is debated.
Mechanisms to uphold standards
Assuming that a new AIS Code is ratified, it is important for that Code to be coupled with an enforcement mechanism. AIS already has in place a set of “AIS
Research Conduct Committee–Process Guidelines” (AIS Process, 2008). In our view, these guidelines provide some amount of reasonable redress when certain prohibited activities–especially those associated with classical plagiarism–occur. We anticipate, however, that a revised AIS Code would address a larger set of activities (e.g., reviewing), and it would be quite unwieldy to expect the existing AIS Process to handle complaints in all of those areas. Further, the existing AIS Process is highly reactive, in that it relies on the editor of an AIS publication receiving a complaint. In its pristine form, the AIS Process provides little in terms of a proactive check-and-balance, especially in areas in which researchers may not be aware of the rules.
We therefore recommend that all the AIS journals embrace a more proactive mechanism, one limited example of which has recently been put into use at MIS
Quarterly . “In the cover letter accompanying the manuscript, authors should confirm that the research described in the paper complies with the ‘Code of Research Conduct’ prepared by the Association for Information Systems.” AIS’s other two major publications – the Journal of the Association for Information Systems (JAIS) and
Communications of the Association for Information Systems (CAIS) do not require such an overt statement on the parts of authors, although they do state that:
15

All prospective authors must adhere to the AIS Code of Research Conduct in their work. Authors who believe their work has been plagiarized or subjected to other scholarly misconduct are urged to consult Guidelines for a Victim. AIS Council's process for dealing with allegations of scholarly misconduct in the Association's journals and proceedings is detailed in AIS Research Conduct Committee Process
Guidelines.
2 (JAIS, 2008)
In our view, MISQ’s approach should be seen as a minimal standard, since it forces the corresponding author to acknowledge an understanding of, and adherence to, the AIS Code. However, we are persuaded that an even higher level of awareness can be effected with an addition embraced by the Journal of Business Ethics (JBE), which expects all authors (not merely the corresponding author) to physically sign a statement declaring that they are personally responsible for the paper’s content, that they have participated fully in the research, and that they have no conflicts of interest associated with the research’s funding. Further, JBE demands that all references to data collection within the manuscript include a notation that certifies IRB procedures were followed.
We believe that combining all these approaches would be helpful, as it would require authors to face head-on some of the issues that seem to disproportionately cause heartache within the research community.
Even with these additional safeguards that would heighten awareness, it is likely that some breaches would occasionally occur; after all, individuals who knowingly engage in academic dishonesty would probably have no qualms about signing a statement
2
This statement is from JAIS. CAIS includes an identical statement but with the phrase “Members of AIS” in lieu of JAIS’s “All prospective authors” and the word “Members” in lieu of the word “Authors” in the second sentence.
16

attesting that they had not done so. For that reason, the enforcement mechanism needs careful consideration. We propose a mechanism to facilitate our discipline to police itself. IS journals should provide an avenue for criticism or correction of published work where (1) authors can publish corrections if errors in prior work are discovered and (2) other researchers can “challenge” articles that examine both ethical and other problems with published works.
CONCLUSION
Frequently, reconsideration of codes of ethics and their enforcement comes as a reaction to a widely publicized incident. We are fortunate that we have not had such an incident in the IS research community in recent years, but our acknowledged plans to engage in more (not less) questionable behavior in the future suggests we should not be complacent. We recommend that our community take the unaccustomed step of proactively examining its ethical standards.
17

APPENDIX 1 – METHODOLOGY
A set of hypothetical scenarios was developed to describe issues highlighted in both the AIS and AOM codes of ethics. This set was also augmented from our personal experiences and those of other senior IS researchers we interviewed. The scenarios can be grouped into two large categories: those related to creating research and those related to the review process.
A questionnaire was developed to measure the extent to which each of the scenarios represented behaviors that were appropriate for IS researchers. The questionnaire presented each hypothetical scenario and asked the respondent to directly assess the appropriateness of the behaviors described. A complete list of scenarios can be found in Appendix 2.
To validate the set of scenarios, we conducted detailed interviews with senior IS and business ethics researchers. These intent of these interviews was to refine both the set of quandaries and the presentation of each scenario while at the same time verify that important ethical quandaries were not omitted from our set. A pilot test of the questionnaire was also conducted to gather feedback on the presentation of the questionnaire. Modifications of the scenarios and the questionnaire were made as a result of feedback during the interview and pilot test stages.
The population for the study was active IS researchers. Active IS researchers were defined as those individuals who had contact information in the AIS faculty directory and had either published in the top 25 IS journals (Lowry et al. 2004) from January 1998 through December 2007 or had attended the International Conference on Information
Systems in the same time period. In total, 4316 active IS researchers were identified to as

members of the population. The questionnaire was administered online; a personalized email was sent to each member of the population inviting them to participate in the study.
It was noted in the e-mail that the survey had been endorsed by president of the AIS and from the editors-in-chief of Information Systems Research , the Journal of AIS , and MIS
Quarterly. A follow-up letter was also mailed to each member of the population.
To ensure the anonymity of all of the participants, none of the respondents was associated with an identifier. Additionally, respondents could use an Internet
“anonymizer” to remove any identifying aspects of requests made of the hosting server.
These safeguards made it virtually impossible to associate an individual response with a respondent.
Of the 4316 researchers in the population, 628 finished the questionnaire, which took 40 minutes on average to complete. Considering returned postal mail and bounced email messages, our conservative estimate is that about 3000 members of the population received the request to participate in the study, giving an effective response rate of 21 percent. Table App1-1 outlines some aggregated demographic information about the respondents.

Table App1-1 – Sample Characteristics
Characteristic Sample
Average
Years since PhD earned
Number of years as reviewer
13
12
Number of publications
More than 100
20 to 100
5 to 19
1 to 4
none
Professional Time
Allocation
Research
Teaching
Service
Other
% of Sample
11.7
40.6
34.2
13.1
0.4
% of Time
40
36
18
6

APPENDIX 2 – SURVEY SCENARIOS
1.
Reese (an assistant professor) and Morgan (a full professor) are colleagues in the same department at a major university. They were almost ready to submit a manuscript they had written with Bailey (a doctoral student) to a respected journal when they learned of Bailey’s decision to discontinue doctoral studies and pursue full ‐ time employment. They knew that Bailey’s new career would not be advanced through academic publication.
Reese and Morgan removed Bailey’s name from the manuscript before submitting it to the journal even though Bailey had made a substantial contribution.
2.
Reese and Morgan were excited about a new research project. They formed hypotheses and designed a project to gather data to test their hypotheses. They considered adding a coauthor. Pat was a valued colleague in their department who would go up for tenure in two years. Pat’s research record was weak and it was clear that Pat could provide no more than token assistance on the project.
Morgan and Reese included Pat on their project solely to improve Pat’s tenure case.
3.
Suppose that Reese and Morgan had not invited Pat and instead considered
Quinn, a senior colleague in their department, and chair of the promotion and tenure committee for their college. Reese would also go up for tenure in the next three years. Reese thought that including Quinn as a coauthor on the paper would provide Quinn an opportunity to see firsthand the quality of Reese’s research, improving the likelihood of a favorable tenure review. Quinn would do little more than read the manuscript and make editorial comments.
Morgan and Reese included Quinn as a coauthor on the paper to improve
Reese’s outlook for a favorable tenure review.
4.
Reese and Morgan held several discussions about authorship order. Reese had done most of the actual work for the project and composed the initial draft.
Morgan had also contributed substantially but felt that they could leverage
Quinn’s substantial name recognition by listing it first.
Morgan and Reese listed Quinn as first author even though Reese had made the most substantial contribution.
5.
Reese and Morgan wanted to complete their study as soon as possible. There were very clear university policies about obtaining formal approval from the
Institutional Review Board (IRB) to collect any data from human subjects. The
IRB at their university routinely granted exemption to studies using surveys similar to the one they would administer.

Confident that such an exemption would have been granted in their case, they collected their data without applying for the IRB approval.
note: the IRB is sometimes called the "Human Subjects Committee"
6.
Suppose the situation were different and the study involved an experiment. The
IRB at their university rarely granted exemptions for experiments and a full review could take as long as 60 days. Reese and Morgan felt that there was virtually no risk to subjects participating in the study.
They conducted their experiment without applying for IRB approval.
7.
In the data collection process, Reese and Morgan asked a research assistant to compile their records into a single database. The subjects had been promised anonymity, but many were friends of the research assistant. Although it was possible for Reese and Morgan to obscure the identity of the subjects in the compilation process, it would have required additional effort. The survey was not of a sensitive nature.
Reese and Morgan turned over their records to the research assistant without obscuring subjects' identities.
8.
While Reese and Morgan were analyzing their data, they became aware of another study examining the same issue. They knew the other researchers were just completing their data collection. Reese and Morgan felt the need to get a manuscript submitted soon and were pleased to find p-values of 0.05 or lower for each hypothesis. Normally, they would have performed tests to verify their data conformed to the statistical model’s assumptions, but they felt that every day that passed before their submission mattered.
They submitted their manuscript without testing the assumptions of their statistical model.
9.
The manuscript was accepted for publication. After it came out in print, Reese reanalyzed their data for a different paper and discovered that the assumptions of the original statistical models were violated. Using another technique to analyze their data, Reese found that many of the relationships in their initial publication were not actually significant. Reese told Morgan of the problems with the original analysis.
They decided to not report the problems with their original analysis.
10.
Suppose that the situation had been slightly different. Reese and Morgan did conduct a thorough analysis of the data prior to drafting the manuscript. They discovered that some of the assumptions of their statistical model had been

violated.
In their manuscript, they did not report that their statistical model’s assumption had been violated.
11.
Suppose Reese and Morgan took a different approach. They were confident that their initial theory and hypotheses were sound even though their hypothesis tests were not quite significant. They knew that if they collected additional data, the increased sample size would have led to support for their hypotheses. To gain significance,
They randomly duplicated records within their current data set to increase their sample size.
12.
Suppose the situation were different. Reese and Morgan were astonished to find that not only did the data not confirm their hypotheses, but the relationships in the data were directly counter to their expectations. In trying to understand their results they searched the literature in a related discipline. It became apparent that the unexpected results were supported by an established theory in the related discipline.
They reformulated their hypotheses and wrote the paper based on the new theoretical insight, showing that the newly constructed hypotheses were supported by the tests.
13.
Suppose the situation were slightly different. Reese and Morgan could not find theoretical support for the unexpected findings. However, their results were compelling enough to suggest a new theory. They were faced with two options:
(1) report the research study as designed, highlighting that the results suggest a new theoretical perspective or (2) to propose a new theory that is consistent with the insights gained from the data collection, adjusting the hypotheses accordingly.
They proposed the new theory and presented the paper as a test of the new theoretical perspective.
14.
Again, suppose that the situation were different. Reese and Morgan found that most of their original hypotheses were supported by significant statistical relationships in the data. Encouraged by their findings, they were anxious to complete a manuscript. Morgan had previously written a paper that used the same theoretical perspective.
To save time, they liberally reused sections from Morgan’s former paper without ever citing it.

15.
In addition to the review that Morgan had already completed, Reese found a well-written doctoral dissertation that had reviewed the literature related to their theoretical perspective. The dissertation cited a number of papers that were not part of the literature they reviewed in the formulation of their hypotheses. It was evident that most of these "new" papers were in harmony with the literature they had already read. Therefore, Reese and Morgan did not feel it necessary to read these papers. However, they felt that including these "new" citations would strengthen their paper.
They cited these "new" papers in their own manuscript without reading the text of the papers.
16.
Reese and Morgan were able to quickly write most of the paper although there was one concept that was particularly difficult to describe. They were aware of a little-known publication that had artfully dealt with exactly what Reese and
Morgan were struggling to communicate.
In writing their manuscript, they used the text from the little-known publication without citing it.
17.
Sidney was a prolific researcher with a reputation for providing thoughtful reviews. On one review, Sidney thought the authors of the manuscript had made a fundamental error in their interpretation of the modeling notation used in their experimental treatments. Sidney was familiar with the modeling notation but was not an expert and sought additional advice from a recognized expert.
Sidney sent the manuscript to an expert on the modeling notation with a note asking about the meaning of particular portions of the experimental treatments.
18.
With the expert’s response, Sidney was comfortable writing the review and recommended that the paper be rejected, which it was. About three months later,
Sidney received a request to review a manuscript on the same topic from another journal and was surprised to see that it was the same paper. Sidney thought, "this will be an easy review to write."
Accepting the referee assignment, Sidney did not disclose having previously reviewed the manuscript to the associate editor (AE).
19.
On a separate occasion, Sidney received a request to review a paper that was on a topic of personal interest. Upon reading the abstract, Sidney realized that this was the manuscript from a conference presentation made earlier that year. Sidney had been at the presentation and was impressed with the research. There was no doubt that the conference presentation and the manuscript were both reports of the same research activity.

Although the journal had a double-blind review policy, Sidney accepted the assignment without reporting to the AE that the author’s identity was known.
note: in a double-blind review, authors' and reviewers' identities are withheld from each other
20.
Sidney completed a very positive review. The next year, while the manuscript was with the authors for revision, Sidney attended the reception of another conference and bumped into the manuscript’s author. The two began to talk about the research presentation Sidney had attended the prior year. Recalling the review, Sidney wanted to let the author know who the reviewer had been, but was uncomfortable just saying it.
Sidney quoted a very supportive line from the review written a few months earlier, leaving the author with no question that Sidney had been a reviewer.
21.
Sidney was surprised to receive a request to review a paper that was authored by a recent coauthor. Sidney had never been asked to review a coauthor’s work and was uncertain what to do. After reading the journal’s review policy and finding nothing about reviewing a coauthor’s work,
Sidney accepted the review assignment and reviewed the manuscript written by a recent coauthor without informing the AE of the coauthor relationship.
22.
While reviewing a blinded manuscript, Sidney was relatively certain of the author’s identity.
Sidney conducted an Internet search to confirm the author’s identity.
23.
Sidney was flattered at the invitation to serve the academic community as an AE of one of the field’s most respected research outlets. Not long after beginning service, Sidney received a request to handle a manuscript that came with the following note from the senior editor (SE): "Sidney, one of the four authors on this paper had a publication with you last year. Under normal circumstances I would not ask you handle this paper, but you are the only one on our editorial board with any experience on this topic. Please do your best with this paper; I'm sure that you can be objective and select a qualified set of reviewers."
Sidney decided to handle the paper.
24.
Later, on a different manuscript, Sidney invited three individuals to serve as reviewers. Each accepted. Subsequently, one reviewer (Dr. Rosen) emailed indicating a family emergency and asked to be released from the review. Sidney agreed and invited another referee. Shortly before the reviews were due, Sidney received a review from Dr. Rosen, making four in all. Two of the reviews,

including the one from Dr. Rosen, concurred with Sidney’s own opinion that the manuscript should be rejected. One recommended acceptance with minor revisions and one recommended major revision,
Sidney decided to ignore the review recommending minor revisions and prepared the AE report using only the other three reviews.
25.
With several years’ experience as an AE, Sidney received a truly outstanding manuscript. With a strong theoretical foundation, comprehensive literature review, novel experimental procedures, solid statistical analysis, and compelling recommendations, Sidney knew the paper would be published but thought "the sooner the better."
In selecting referees, Sidney considered only those who had a history of being "easy" on authors in order to reduce the time before the paper’s publication.
26.
A short time later, Sidney received a manuscript that had little chance of being improved enough to meet the journal’s standard for publication; however, it was not a clear candidate for a desk rejection. In selecting referees,
Sidney considered only those who had a history of being "hard" on authors to increase the probability of the paper receiving negative reviews.
27.
Many years later, Sidney received an assignment to handle a manuscript authored by Robin Albinson. Sidney remembered the many painful interactions with Dr. Albinson back in that first doctoral seminar and about hearing (second hand) that Dr. Albinson had tried to have Sidney dismissed from the doctoral program after the first-year examinations. Although more than 15 years had passed, Sidney still felt resentment over the events of that first year and wondered if an old grudge would prevent a fair disposition of the manuscript now under consideration.
Sidney decided to accept the assignment to handle the manuscript.
28.
Once, Sidney reluctantly accepted an assignment to handle a manuscript. The topic was a familiar one, but Sidney was no expert and had difficulty selecting appropriate reviewers. Weeks turned into months and the manuscript still had not been sent out for review. Finally, referees were selected and the manuscript went out for review. However, one referee was late in returning a report. Again many weeks elapsed. Several times Sidney thought of sending a message to the delinquent referee; however, other urgencies always seemed to take precedence.
At last the final review arrived and after another delay, Sidney wrote the AE report and the paper was rejected.
Through Sidney’s neglect, the review process took over eight months—twice

the time stated by the journal’s review policy.
29.
Although the responsibilities of being an AE took considerable time, Sidney continued collaborating with several coauthors as they moved their own research forward. Sidney and one coauthor were pleased with a paper they had recently submitted to a respected outlet. Not long after submission, Sidney was assigned to handle a paper on the same topic. Upon reading the manuscript, worry set in.
Not only did the paper address the same topic, but it was clearly superior in almost every regard. Sidney was certain that the manuscript would be published and feared that if were published too soon it would reduce the likelihood that
Sidney’s own work would be accepted.
In handling the manuscript, Sidney extended the review process by recommending a "major revision" of the manuscript knowing that a
"conditional acceptance" was more appropriate.
For each of the 29 scenarios, the survey requested respondents to make the following six judgments. In each, the “behavior” references refers to the bolded portion of the scenarios shown in Appendix 1. Each judgment is made on a seven-point scale.
1.
Assessment of the inappropriateness (1=clearly appropriate, 4=neutral, 7=very inappropriate)
Judgments two through six are frequency estimates (1=never, 7=often).
2.
I have felt pressure from others to engage in similar behavior
3.
I have engaged in similar behavior
4.
Colleagues in my discipline with whose behavior I am personally familiar engage in similar behavior
5.
In general, I believe that researchers in my discipline engage in similar behavior
6.
When facing similar situations in the future, I would likely engage in similar behavior

APPENDIX 3 – STATISTICAL SUMMARY
Scenario
Drop a colleague as an author
Add a coauthor to improve coauthor's chance at tenure
Add a coauthor to improve own chance at tenure
Author order other than contribution
Skip IRB for survey
Skip IRB for experiment
Violate anonymity of subjects
Not test statistical model assumptions
Not report discovery of error subsequent to publication
Not report violations of a statistical model assumptions
Randomly duplicate data to increase sample size
Reformulate hypotheses based on existing literature
Reformulate hypothesis, show results as supporting new theory
Self plagiarism
Citing without reading
Classical plagiarism
Referee circulating a manuscript outside of the review process
Re-reviewing a manuscript
Double blind review where author's identity is known
An individual revealing that he or she was a reviewer
Reviewing for a recent coauthor
Searching for an author's identity in a double-blind review
Handling a paper for a recent coauthor
Excluding a review from the AE report
Choosing easy reviewers
Choosing hard reviewers
Handling a paper for an enemy
Delaying a review through neglect
Delaying a review for self interest
Means Standard Deviations
M1 M2 M3 M4 M5 M6 M1 M2 M3 M4 M5 M6
6.64 1.57 1.14 2.25 2.84 1.18 0.83 1.24 0.53 1.56 1.42 0.67
4.92 2.82 2.42 3.61 3.98 2.69 1.54 1.89 1.67 1.86 1.69 1.65
5.40 2.38 1.80 3.17 3.74 2.09 1.50 1.82 1.43 1.97 1.83 1.49
5.80 2.42 1.84 3.18 3.78 1.88 1.52 1.82 1.43 1.90 1.83 1.34
5.55 2.33 2.02 2.90 3.37 2.13 1.58 1.81 1.62 1.94 1.86 1.63
6.16 1.69 1.44 2.17 2.69 1.63 1.37 1.37 1.15 1.63 1.69 1.31
6.02 1.57 1.47 2.14 2.80 1.65 1.28 1.24 1.08 1.51 1.63 1.27
5.32 2.12 1.69 2.78 3.60 1.98 1.50 1.64 1.24 1.81 1.78 1.37
6.15 1.68 1.34 2.27 3.23 1.58 1.22 1.36 0.85 1.65 1.75 1.08
6.54 1.62 1.30 2.23 3.06 1.37 0.93 1.34 0.81 1.64 1.71 0.90
6.77 1.31 1.11 1.72 2.30 1.14 0.79 0.99 0.52 1.33 1.51 0.61
3.29 2.75 2.83 3.76 4.49 3.62 1.93 1.99 1.95 2.00 1.83 2.10
3.61 2.55 2.49 3.44 4.27 3.31 2.02 1.94 1.88 2.00 1.80 2.01
5.78 2.26 1.92 3.16 3.82 1.95 1.51 1.74 1.41 1.96 1.92 1.44
5.33 2.63 2.63 3.66 4.39 2.49 1.49 1.87 1.60 1.93 1.86 1.58
6.83 1.34 1.14 1.88 2.58 1.15 0.78 0.98 0.56 1.46 1.56 0.64
3.05 2.06 2.48 3.10 3.70 3.45 2.04 1.64 1.80 1.87 1.77 2.05
5.75 1.48 1.37 2.15 2.92 1.60 1.52 1.17 1.04 1.63 1.72 1.28
4.66 1.85 2.12 2.99 3.80 2.56 1.73 1.54 1.67 1.87 1.81 1.82
5.53 1.60 1.58 2.56 3.23 1.81 1.58 1.24 1.18 1.72 1.70 1.29
5.35 1.77 1.65 2.65 3.46 2.03 1.73 1.49 1.33 1.83 1.79 1.53
4.40 1.85 2.64 3.18 3.95 2.85 1.66 1.51 1.83 1.93 1.82 1.80
3.11 2.08 2.08 3.20 4.24 3.65 1.83 1.70 1.71 1.98 1.78 1.90
5.59 1.52 1.46 2.31 3.06 1.89 1.74 1.21 1.18 1.72 1.82 1.55
5.26 1.65 1.52 2.50 3.51 2.14 1.62 1.33 1.19 1.77 1.85 1.54
5.66 1.59 1.42 2.38 3.39 1.81 1.50 1.27 1.08 1.78 1.84 1.33
5.03 1.44 1.36 2.21 3.22 2.14 1.75 1.11 0.95 1.61 1.67 1.58
5.89 1.81 1.88 3.22 4.34 1.94 1.25 1.53 1.41 1.96 1.81 1.26
6.76 1.37 1.14 1.97 3.07 1.25 0.80 1.09 0.57 1.56 1.80 0.73

REFERENCES
Academy of Management Revised Code of Ethics. 2008. http://www.AoMonline.org/AoM.asp?id=14&page_id=235 , accessed on August 18,
2008.
Association for Information Systems Code of Professional Conduct (AIS Code). 2008. http://home.aisnet.org/displaycommon.cfm?an=1&subarticlenbr=15 , accessed on August
18, 2008.
Association for Information Systems AIS Research Conduct Committee – Process
Guidelines (AIS Process). 2008. http://home.aisnet.org/displaycommon.cfm?an=1&subarticlenbr=12 , accessed August 19,
2008.
Frankena, W.K. Ethics , Prentice-Hall, Englewood Cliffs, NJ, 1973.
Gragg, C.I. “Because Wisdom Can’t be Told,” Harvard Business School note # 9-451-
005, 1940.
Journal of the AIS (JAIS), “Code of Research Conduct,” in “JAIS Policy,” http://jais.aisnet.org/policy.asp
, accessed August 19, 2008.
Kock, Ned. “A Case of Academic Plagiarism,” Communications of the ACM , vol. 42, no.
7, July 1999, pp. 96-104.
Kock, Ned and Robert Davison. “Dealing With Plagiarism in the Information Systems
Research Community: A Look at Factors That Drive Plagiarism and Ways to Address
Them,” MIS Quarterly , vol. 27, no. 4, December 2003, pp. 511-532.
Laudon, K.C. “Ethical Concepts and Information Technology,” Communications of the
ACM (38:12), 1995, pp. 33-39.
Lowry, P. B., Romans, D., & Curtis, A. Global journal prestige and supporting disciplines: A scientometric study of information systems journals, Journal of the
Association for Information Systems (5:2), 2004, pp. 29-75.
Management Information Systems Quarterly (MISQ), “Compliance with AIS Code of
Research Conduct and Sample Cover Letter,” MISQ Roadmap, http://www.misq.org/roadmap/standards.html
, accessed August 19, 2008.
Smith, H.J. and Hasnas, J. “Ethics and Information Systems: The Corporate Domain,”
MIS Quarterly (23: 1), 1999, pp. 109-127.

Low
Low
Low
Low
Low
Low
Low
Low
Research
Emphasis*
Low
Low
Low
Low
Future
Future
Future
Future
Future
Future
Future
Future
Future
Future
Future
Future
Future
Future
Past &
Future t
Future
Future
Future
Future
Future
Future
Future
Future
Future
Future
Future
Future
Future
Future
Future
Childhood
Education*
AIS
Membership*
Items
Drop a colleague as an author
Non NAmer Non Members Add a coauthor to improve coauthor's chance at tenure
Non NAmer Non Members Add a coauthor to improve own chance at tenure
Non NAmer Non Members Author order other than contribution
N America Skip IRB for survey
N America Skip IRB for experiment
Violate anonymity of subjects
Non NAmer Non Members Not test statistical model assumptions
Not report discovery of error subsequent to publication
Not report violations of a statistical model assumptions
N America Randomly duplicate data to increase sample size
AIS Members Reformulate hypotheses based on existing literature
AIS Members Reformulate hypothesis, show results as supporting new theory
Self plagiarism
AIS Members Citing without reading
N America Classical plagiarism
Referee circulating a manuscript outside of the review process
N America AIS Members Re-reviewing a manuscript
N America AIS Members Double blind review where author's identity is known
An individual revealing that he or she was a reviewer
AIS Members Reviewing for a recent coauthor
N America AIS Members Searching for an author's identity in a double-blind review
Handling a paper for a recent coauthor
Excluding a review from the AE report
Choosing easy reviewers
N America
AIS Members Choosing hard reviewers
Handling a paper for an enemy
AIS Members Delaying a review through neglect
Delaying a review for self interest
*- Examine differences between groups for perceived inappropriateness of the behavior; noted group has higher mean (see item t as more inappropriate)
– Past & Future refer to extent of engagement and expected future engagement in the described behavior
Table 1 – Segmentation Analysis of Ethical Scenarios
Explicitness of the AIS Code
2 - The AIS Code Directly Addresses the Behavior
Inappropriateness Past Engagement Future Engagement
5.88 * 1.66 * 1.86 t
1 - The AIS Code Vaguely Addresses the Behavior 5.76 * 1.54 * 1.85 t
0 - The AIS Code is Silent on the Behavior 4.82 * 2.00 * 2.44 *
* Means within columns significantly different at p < .01 t Means are not statistically different, but are statistically different from other means in the column at p < .01
Table 2 – Does the AIS Code Influence Attitudes, Behaviors, and Intentions?