Searching for academic honesty

Topics/tags: Miscellaneous, the Web, Grinnell, rambly

The other day, one of my colleagues sent me the following message.

Could you write a musing about why when I search Academic Honesty on the college’s website, I can not find the current policy (2018-19), but your musing is the second hit, and the majority of the remaining hits are CS pages?

That’s an interesting question. I don’t have a great answer, but that’s never stopped me from musing about an issue before.

First, let’s see what comes up when you search for Academic Honesty [1] on Grinnell’s Web site.

• The first result is the 2017-2018 Academic Honesty Handbook.
• The second result is one of my musings on academic honesty [2].
• The third result is the 2013-2014 Academic Honesty Handbook. No, I’m not sure why we don’t see the intervening handbooks.
• The fourth result is my academic honesty policy from the Fall 2015 session of CSC 151.
• The fifth result is the Academic Advising Office’s list of publications.
• The sixth result is the CS department’s policy on academic honesty [3].
• The seventh result is a talk by Henry Walker on Academic Honesty. Henry spent many years on CAS and often gets asked to talk to different groups of students about academic honesty issues.
• The eight result appears to be a slide from that talk.
• The ninth result is another of my musings on academic honesty.
• The tenth result is another slide from one of Henry’s talks.

That’s the first page of results. And we all know that no one goes beyond the first page of search results.

Interestingly, you see a somewhat different set of results if you search for Academic Honesty Grinnell [4] on Google.

• Once again, the first result is the 2017-2018 Academic Honesty Handbook.
• This time, the second result is the Academic Policies and Procedures page in the 2018-2019 Student Handbook.
• The third result is my academic honesty policy from the spring 2016 offering of CSC 151. No, I don’t know why the different searches choose different courses from which to extract the mostly identical policies.
• The fourth result is the CS department’s policy on academic honesty. It appears we’re moving up in the world.
• The fifth result is the 2013-2014 Academic Honesty Handbook. That’s moving down a bit.
• The sixth result is a page on Honesty in Academic Work from the 2018-19 Student Handbook.
• It’s not until the seventh result that we reach one of my musings on academic honesty.
• The eighth result is the Academic Policies page from the 2016-2017 College Catalog. Why isn’t it from the current catalog? That’s a good question. I’m also intrigued to see that the result highlights academic dishonesty.
• Once again, the ninth result is another of my musings on academic honesty.
  
• The tenth result is a guide to creating an inclusive syllabus.

So much to ponder. So little time.

Let’s start with how search engines often rank pages. At least in the early days, the ranking of a page depended on the words that appeared in the page (or at least the words relevant to that search) and the reputation of the page. The reputation of a page was computed based on the number of links to that page and the reputation of the linking pages [5]. The algorithm has since been updated in a variety of ways because folks looked for ways to game the system. But the number of incoming links and reputation are still significant factors.

That’s likely one of the reasons that many CS pages appear on searches. We’ve had a Web site for almost as long as there has been a Web, we don’t tend to move our pages, and we tend to write interlinked pages. The reputation of older pages may then influence the reputation of newer pages, particularly if the older pages link (indirectly) to the newer pages.

There are techniques for Search Engine Optimization (SEO, in the annoying TLA parlance). But I’ve never worried about them, and I don’t think anyone else in the department does, either.

What about the non-CS pages? I note that many of the most relevant results appear in the general Google search rather than Custom Grinnell search. As far as I can tell, it appears that the College’s custom search does not index catalog.grinnell.edu. I’m not sure why not. It also doesn’t index grinco.sharepoint.com. That’s because we hide that content behind a password wall. We were promised a non-protected Sharepoint site a few years ago, but that has yet to become a reality.

Why did Henry’s pages show up in the Custom Grinnell search but not the general Google search? I have no idea.

What’s next? I believe my colleague was looking for the latest version of the academic honesty handbook. So I did what I normally do when looking for a variant of a document I can find. I loaded last year’s handbook, which was conveniently the first result, and substituted 2018-19 for 2017-18. Unfortunately, that did not work.

Next, I tried a different search. Since the subtitle of the Academic Honesty Handbook is Scholarly Integrity, Collaboration, and the Ethical Use of Sources, I searched for Scholarly Integrity. The first result was the Academic Advising Publications page with the text Academic Honesty Booklet (PDF file). So I went to that page and, lo and behold, there was a link to the current Academic Honesty Handbook. It appears that it’s been reformatted, but the content is much the same. Should it bother me that it contains sections I wrote [6], but I don’t appear explicitly in the acknowledgements? Probably not.

Have I answered all my colleague’s questions? Let’s see. Why didn’t you find the current policy? Because Grinnell seems not to index the College Catalog and that’s where you can find the policy. Why couldn’t you find the current booklet? Because it doesn’t appear immediately in search results and you probably were not inclined to do a broader search [7]. Why do CS pages appear high in these search results? Because we’ve built a good reputation with the Google search engine. Did I catch them all

There may be one other, which was implicit in the original message. Why do computer scientists write so much about academic honesty? I expect there are a variety of reasons. One is that we seem to encounter more issues of academic (dis)honesty than other faculty. That’s not unique to Grinnell. For example, Harvard’s CS 50 course reports more students for academic dishonesty than any other course there, and that’s with policies that let students withdraw a homework up to 72 hours after submitting it [8]. Do students cheat more in our classes? I don’t know. But it may be easier to detect copying in CS than in some other disciplines [9]. It may also be that students don’t realize that copying parts of a program can constitute dishonesty; after all, they often copy formulae in their other science classes. We write about academic honesty to help clarify these issues for our students [10].

There are, of course, students who know that they are cheating. I recall one instance from early in my career [11] in which student P helped student Q on a take-home exam by taking P’s code and renaming all of the variables and giving the new program to Q. Rebelsky won’t notice. But I did. Then I got my first experience with idiotic legalisms. The Academic Honesty Committee said, There’s nothing in your syllabus that says that it’s dishonest to give your work to another student. That’s another reason I try to make my policies clear.

As computer scientists, we also care about academic honesty because it has an impact on our students as professionals [12]. If you copy code in industry and don’t obtain permission, you put your employer at legal risk. I would also venture to guess that those who succeed by uncredited copying are less likely to produce software that we should trust.

It’s likely that there’s also something about the character of this department that leads us to write a lot about academic honesty. As I said, Henry Walker served for many terms on CAS. I got involved in parts of the handbook during some part of my Grinnell career. So we may write more about these issues than CS faculty at other institutions [14]. Others have likely followed our lead.

That’s about all I can think of as an answer to the original question. I hope it suffices.

Postscript: Note that I do not make it a regular practice to respond to requests to muse about particular issues. But I value this colleague, and I found it a compelling question. I even learned a few things along the way.

[1] Without quotation marks.

[2] Did I really muse about academic honesty on January 1? I must have been quite upset.

[3] There’s a long story behind that policy. Maybe I’ll write it at some point. Suffice to say, at some point, the Committee on Academic Standing suggested to the CS Department that we should write our own policy statement. Like most CS department documents, this one was authored collaboratively.

[4] Again, without the quotation marks.

[5] Yes, that’s a recursive formulation with no clear base case. I believe you can compute a fixpoint of that function, or at least an estimate thereof.

[6] I pushed hard that the handbook better clarify that there are multiple kinds of academic honesty beyond plagiarism. It looks like at least three of the questions in the FAQ are things I wrote.

[7] Or maybe you did, and just wanted to know why my musings appear high in the list of results.

[8] There’s some evidence that a lot of cheating happens when students are tired late at night. The 72-hour policy lets them reflect on their decision when they are more rested and avoid penalty.

[9] Some schools that run plagiarism detectors for code. We don’t. Rather, we tend to note copying because of unexpected similarities in atypical answers.

[10] And, at times, for colleagues in other disciplines.

[11] Not at Grinnell.

[12] Not all CS majors become computing professionals, but many do.

[14] Those who’ve dealt with Henry and me know that we both have a tendency [15] to write a lot.

[15] Grammarly says to replace have a tendency with tend. But I think leaving the former in helps make my point.

Verson 1.0 of 2019-01-16.