The following is from Draft 2 of the ACM Policy on Complaint Process Disclosure.
If you wish to contribute to this section: Reply to this post with your suggested re-wording, justification for your proposed changes, and an indication of your support or not.
3 Reporting and reflecting
The ACM shall collect and make publicly available aggregate data summarizing the number of complaints filed, the category of complaints filed, and within each category, the number dismissed without investigation, the number investigated with no violation found, the number of minor violations, the number with more serious violations, and the number still open.
For complaints at a level above a minor violation ACM will make publicly available information that includes the nature of the offending action, the date or general time frame of the offending action, the venue, and the policy violated.
ACM will inform the community of the results of applying this policy before June 30, 2027. The report will evaluate the effects of this policy and recommend changes that may need to be made.
Just saying that the data is collected and made publicly available is vague. The timeline for making the data public should be specified, at least at a minimum. I would recommend adding “at least annually”, which would align with the ACM Annual Report. More frequent publication of data may be available elsewhere, but it sets a baseline.
The date of June 30, 2027 is odd, considering that it’s 4 years away. Is there a reason this date was selected?
I am in favor of a policy in which adjudicated decisions of the E&P committee in which fault was found are publicized in a de-identified manner. As a member of the E&P Committee, I have been continually frustrated that there is no public record of the cases in which rules have been found to be violated.
This is not about naming and shaming. This is about using these cases as teachable moments. It is also about creating a body of case law that presents what the E&P Committee did in difficult situations.
I understand that some cases cannot be de-identified. Typically, those are high-profile cases. In those situations, I favor having the full decision made available in which the responsible parties are named.
The date of June 30, 2027 is odd, considering that it’s 4 years away. Is there a reason this date was selected?
Given that this policy is a departure from current practice in a significant way, we want to ensure that its effects are given careful consideration. After near the end of four years in operation, we are hopeful that its impact can be evaluated, and if need be, adjusted.
There is value in demonstrating the sorts of decisions both Ethics and Plagiarism and the Committee on Professional Ethics have made. It demonstrates ACM’s commitment to upholding high standards and, as you point out, sharing those “teachable moments.”
The intention is for the policy to be voted on by Council in June. The evaluation of its impact takes place in four years should the policy be adopted.
Four years seems like an awfully long time to evaluate the impact of the policy. Is there a written plan as to why four years are needed, and what will happen during each year?
I think Simson’s point here is excellent and as a long standing member of the Ethics and Plagiarism Committee, he has substantial expertise on which to base his recommendations.
Where there are patterns of abuse and misconduct, there are substantial benefits to sharing those patterns in a de-identified way so that the community knows what practices to watch out for vs. just what people. For example, in the actual case highlighted in the side bar of this article, we learned (sadly slowly) about individuals were bidding on each other’s papers and sharing information. When the community knows about the holes in processes, they can change practices to thwart them.
When bad practices are not discussed more openly, the community remains more vulnerable not just to the same bad actors but to other bad actors who might exploit the same holes.
Simson said it perfectly, This does not have to be about naming and shaming. It can be about teachable moments where the community learns to better protect itself from abuse that they did not even know was occurring!
Harm prevention is not just about the sanctioning people. It also about the community evolving to understand what practices it need to guard against.
Four years seems like an awfully long time to evaluate the impact of the policy.
In our estimation it will take a year to get this implemented in a complete way, years 2 and 3 will provide the bulk of the information, with some of it coming during year 4 while the report is being finalized.
Is there a written plan as to why four years are needed, and what will happen during each year?
No plan yet. Such a plan should reflect the actual policy that Council adopts. We expect the draft to change as a result of this process.
I agree that this is vague. In addition to describing the issue, ACM should also publicly release information about its actions after the investigation is concluded, like remediation or sanctions (ok to anonymize).
For complaints at a level above a minor violation, ACM will make publicly available information that includes the nature of the offending action, the date or general time frame of the offending action, the venue, and the policy violated.
In particular, we feel this will de-identify the complainant and therefore lead to under-reporting of such violations. The fine granularity of the venue and the nature of the action are particularly problematic. In order to decide if this level of granularity is acceptable, we need to know the categorizations in place and the number of incidents assigned to each categorization. For example, it would potentially be okay to release data on total harassment claims at all ACM conferences in a given year but not harassment claims at say EC’23. We would like to emphasize our concern is about complaints where there is stigma for the victim such as harassment. We are fine applying this policy to violations such as plagiarism.