Editorial

This issue contains two papers on testing of special-purpose software. Systematic testing of actor systems, by Elvira Albert, Puri Arenas, and Miguel Gómez-Zamalloa, presents ideas for testing concurrent systems. In particular, instead of testing all possible interleavings, the technique prunes the states that are explored to reduce unneeded non-determinism. (Recommended by Yves Le Traon.) Verifying OSEK/VDX automotive applications: A Spin-based model checking approach, by Haitao Zhang, Guoqiang Li, Zhuo Cheng, and Jinyun Xue, presents a use of model checking to verify automotive software. (Recommended by Shaoying Liu.)

Proper References is a Matter of Scholarship, Ethics, and Courtesy

Conferences often have associated workshops, some quite successful. For example, the International Conference on Software Testing, Verification, and Validation hosts several workshops every year. Unfortunately, this habit has encouraged the proliferation of a type of bad scholarship. Consider the following workshop paper:

The publisher posts workshop papers online, and as an added convenience, includes the reference information. Except they often get the reference information wrong!

This paper is listed on IEEE Explorer (http://ieeexplore.ieee.org/document/7899049/) as:

The problem, of course, is that this does not include the workshop name. There is not even a conference named ICSTW!

Of course, this is the publisher’s mistake. Several program chairs of major conferences have asked IEEE to stop doing this, without success. They are publishers, not scholars, so maybe they do not understand. But scientists who truly care about scholarship cannot repeat their mistake. We have to correct it. And sometimes it takes work—for one paper, I had to contact the authors to ask which workshop their paper was in.

Getting references correct is proper scholarship and should be a matter of pride. It also matters to your reputation—if readers and reviewers cannot trust you to get the reference right, will they trust you to get the data right? Experimental results are much harder to check.

Sometimes, this becomes a matter of ethics. You certainly knew in which workshop your paper appeared. If you mis-reference your own paper, it seems as if you are trying to falsely claim credit for a paper in the main conference (a more prestigious publication).

A couple of years ago I found an improper reference on a CV. The paper was in ICSE’s New Ideas and Emerging Research track (NIER), which has some interesting ideas, but NIER papers are not full papers, not fully reviewed, and usually preliminary. The CV listed the NIER track paper as simply “International Conference on Software Engineering,” as if it was in the main track. Opinions varied from honest mistake to bad advice to outright fraud. Regardless of the root cause, we immediately disqualified the application. Don’t take that risk!

My last point is about listing authors in references. Here is a subtle suggestion: Never, NEVER, NEVER use “et al.” in the reference list. (Okay, that’s not really very subtle.) Using “et al.” is fine as a shortcut in the text, but using “et al.” in the reference list is disrespectful to the other authors and hides information that may be relevant to readers. The missing authors worked on the paper and deserve credit. Sometimes they deserve more credit than the first author, for example, if the authors’ policy is to list names in alphabetical order.

Long ago, a colleague told me he rejected a paper because his name was omitted in a reference. I cannot accept that as proper behavior, but from an author’s perspective, why give an unreasonable reviewer such an opening?

Good scholarship is about more than generating good technical results. We are a community of scholars, and to be accepted and respected as a member of the community, we all need to be good colleagues. And that means getting the references right.

Jeff Offutt
George Mason University
offutt@gmu.edu
26 March 2018