Not such a wise OWL

Print Friendly, PDF & Email

It came as a bit of a shock, a press release declaring The Purdue University Online Writing Lab and Chegg Partner to Make World-Class Writing Education Tools More Accessible.

Purdue’s Online Writing Lab, often referred to as “the OWL at Purdue.” is a much-respected service, providing advice on academic writing in all its aspects, most especially for its comprehensive guidance on the formatting of MLA, APA and Chicago references. .  For many, it is the number-one go-to guide.

I have to admit, the OWL at Purdue is not my number-one source.  For my own referencing queries, I go directly to the official publications,  the MLA Handbook, the APA Publication Manual or Turabian’s A Manual for Writers of Research Papers, Theses and Dissertations, a version of the 1144-page Chicago Manual of Style adapted for use in colleges and schools.  If I cannot easily find what I am looking for, I will try the official Q&A sites: Ask the MLA at the MLA Style Center, the APA Style Blog and the Chicago Manual of Style Online Q&A are great supplements to the guides themselves. I also use Noodletools.

But there are those and they are many who prefer and recommend the OWL at Purdue.  Fine.

There are of course, also those who prefer technological solutions such as EasyBib, BibMe, Cite This For Me, Citation Machine, CiteFast, WorksCited4U and Noodletools (and there are many many more) over the published manuals, but there we go.

Note, I hesitate whether to include Noodletools in this list because it is in a class of its own.  Among other considerations,

  1. the designers of Noodletools go to great pains to ensure that the advice they give and the references generated are accurate and up-to-date,
  2. Noodletools offers far more than just a reference generator (including note-cards,  organising tools and collaboration opportunities),
  3. its pop-up help and advice messages really are helpful, promoting accuracy,  consistency and completeness, and
  4. it is a superb learning tool throughout: students get to know what they are doing, will learn to be independent, will learn the elements which make up a reference and more readily see the patterns, and will be able to stand on their own feet even when asked to use a referencing style other than the Big Three.

[Disclosure of possible bias: I have worked with Debbie Abilock, the founder of Noodletools, on a number of occasions. I hold Noodletools in the highest regard, not out of friendship or collegiality, but out of firm conviction that it really is the best of its type.]

These considerations are not necessarily true of other online reference generators.  Most generators (including Noodletools) offer a selection of forms one must choose according to the type of source (book, magazine, newspaper, online newspaper, web page etc etc); then you manually complete the form with the information available in the source being referenced: author/s, title of source, date of publication, other publishing details, URL and so on.  The software algorithm then produces a reference in the chosen style.

One issue is that, in many reference generators,  the forms are not necessarily accurate. They may, for instance, miss asking for elements which really should be included (or request details which are not needed).  Another is that help, advice or tips are often sparse (and may even be inaccurate), leaving users to guess at (for instance) capitalization and punctuation or dates and other elements required.  There is often inconsistency between the results generated by different reference generators.  Worryingly, there are often internal inconsistencies thrown up by individual reference generators.

(This is where Noodletools scores. As noted, the designers go to great lengths to ensure accuracy in terms of information to be entered for each type of source, to assist the user in entering the information correctly, and in ensuring the software accurately generates the references based on information input. I am not so sure about other referencing generators.

(This too is where the OWL at Purdue scores. As well as its templates for referencing in the various styles, it offers videos. tutorials and other resources to help students learn what is needed and why it is needed. The OWL is about writing, not about getting the formatting right.)

Many online reference generators offer the opportunity to auto-cite sources:  just enter a URL or ISBN or title and the auto-cite feature will come up with one or more suggestions, and generate a reference in the requested style.  Auto-citation is notoriously inaccurate.  Much depends on the accuracy of the metadata of the requested source – often incomplete or given the wrong label or entered in the wrong data field – and then on the ability of the user to determine if elements are missing or in the wrong place in the generated reference, if the punctuation is correct and so on.

If the user is reliant on reference generators and has not learned what a correctly-formatted reference (in the chosen system) looks like or should include, then there is no way the user will know if the generated reference is wrong or has missing elements.  Knowing what a correct reference looks like is of course not essential if one has a good template or examples to hand to check the accuracy of the generated reference. This is where the appropriate referencing manual/ style guide or the OWL at Purdue or any other example sheet is helpful.

Auto-citation is not necessarily a problem in itself.  Too often, however, users use auto-citation tools and trust implicitly in their output. They do NOT check for accuracy or completeness.  They want ease and speed. Checking for accuracy slows them down so mistakes are unnoticed, are accepted.

None of this is new.  I have critiqued and criticized reference generators and especially auto-citation several times in previous posts, including A gift that kept on giving…, Not so easy does it, Getting it wrong…, and Double-dipping.

In one of these posts,  By any other brand-name, not so sweet? I expressed dismay at the Chegg company’s take-over of many reference generators.

Chegg makes me feel uneasy. It advertises “24/7 homework help,” online tutors and other study help and solutions manuals (solutions to problems posed in textbooks).

The Chegg stable also houses several reference generators, although only EasyBib is mentioned on its website.

The help offered and the use made of the service may be dubious, as evidenced by comments made in the Chegg Homework group on Twitter or the complaints made on ConsumerAffairs.com – many other review sites are available.  Apart from those hosted by Chegg, they do tend to be unfavorable.

What’s more, a recent trend is for reference generators to offer more than reference generation – including proofreading (spelling and grammar checks) and text-matching services (sometimes but erroneously called ” plagiarism detection”).

Anyone who has qualms about uploading work to Chegg should be wary (we all should be wary).  Note, for instance, a note of reassurance on Citation Machine’s so-called smart proofreader:

Don’t worry, your writing won’t be searchable publicly.

Maybe your writing won’t be “searchable publicly,” but you are still giving it away.  Chegg’s Terms of Use (covering all its products) should give pause:

User Content and Activities

When you submit, post, upload, embed, display, communicate, link to, email or otherwise distribute or publish any review, problem, suggestion, idea, solution, question, answer, class notes, course outline bibliographic and citation information comment, testimonial, feedback, message, image, video, text, profile data or other material (“User Content“) to Chegg, any Chegg employee or contractor, or a Chegg Website, you grant Chegg and our affiliates, licensees, distributors, agents, representatives and other entities or individuals authorized by Chegg, a non-exclusive, worldwide, perpetual, unlimited, irrevocable, royalty-free, fully sublicensable (through multiple tiers) and fully transferable right to exercise any and all copyright, trademark, publicity, and database rights you have in the content, in any media known now or in the future, and to make, use, reproduce, copy, display, publish, exhibit, distribute, modify, sell, offer for sale, create derivative works based upon and otherwise use the User Content.

Note that we may create, facilitate or display social advertisements, whereby your name, profile and photo may be used to advertise products and services to your network based on your use of the Services and your interactions with Chegg. You agree that Chegg may use your name and profile picture in connection with social ads to advertise products and services to your network based on your use of the Services and your interactions with Chegg and third parties through the Services.

You could be giving Chegg the right to use your work in any way they wish.  By uploading, you are giving away your essays. By using the site, you are giving Chegg the right to use your personal data in any way they wish, including “your name, profile and photo.” Is this wise?

A note in a later section of the Terms of Use sounds remarkably close to disclaimer statements made by many companies which sell pre-written and custom-written essays:

Homework Help, Study Guide and Note Usage

Chegg’s homework help, study and note services are available merely as informational and study aids and should not be considered substitutes for applicable coursework, homework, class and lecture requirements, assignments and related materials. In using the Services, you specifically agree not to use, claim or submit as your own any portion of the help materials. Chegg does not guarantee the accuracy or quality of answers or other study material that appear on the Services, some of which may be posted by other users. You further agree the Services may present information that is incorrect or inconsistent when compared to similar content and materials, including solutions and their methodologies, provided or preferred by publishers of applicable problems or instructors of applicable courses.

I really do have strong reservations about Chegg and its various off-shoots.

So I am saddened by the news that the OWL at Purdue is teaming up with Chegg. The OWL looks set to tarnish its independence and its reputation.

I find several aspects of this partnership disturbing.

That press release The Purdue University Online Writing Lab and Chegg Partner to Make World-Class Writing Education Tools More Accessible includes the declaration:

Chegg’s writing tools will now integrate OWL’s rules and standards in order to teach students to become better writers. Students will upload their documents and receive AI-enabled instant feedback on grammar, spelling, and how to write more effectively as well as access resources from the OWL on any aspect of writing and at any stage of its development. In this dynamic learning experience, students will receive immediate feedback, with deep context and rich examples.

Users of the now defunct RefMe reference generator know only too well that when Chegg bought out RefMe and merged it with CiteThisForMe, it did not raise the quality of CiteThisForMe.  It did horrify those who had used RefMe; quality suffered. Change is not always for the better.  Look out, OWL!

Again disturbingly, an article on the partnership in Inside Higher Education The Wrong Partnership? carried statements attributed to Harry Denny, director of OWL, including this one:

OWL will advise Chegg on writing instruction and help to develop the company’s AI-powered writing improvement tools. Chegg will license OWL’s writing tips and place advertising on its website, helping OWL to monetize its free content. Purdue students will also be given the opportunity to apply for paid internships at Chegg, said Denny. OWL anticipates a six-figure revenue from the partnership, he said.

All kinds of alarm signals are sounding, “helping OWL to monetize its free content” for one. On the internet, too often we pay for free content with our data and our privacy.   There is certainly money and more to be made from our data – just look at Facebook.

It could well be a win-win situation for Chegg, for OWL users are invited to use Chegg services:  OWL is now promoting (at least) one particular Chegg product, Citation Machine.

Above: the OWL at Purdue as on 2 February 2019 (courtesy the Internet Archive).
Below: the OWL at Purdue as on 3 March 2019 (courtesy the Internet Archive).

Is Citation Machine worth trying? It seems prone to the same mistakes as other trials I have made with my go-to citation requests – but it does make those mistakes faster.  Is that a plus?  (That’s a rhetorical question – don’t even try to answer that.). Here are the suggestions for one of my favorite try-this auto-cites, in MLA, APA and Chicago and respectively.

The article, published in The Journal for Historical Review (Vol. 6, No. 1) in Spring 1986, pages 9 ff. and posted on the Institute for Historical Review web-site is by John Bennett; it carries the title “Orwell’s 1984: Was Orwell Right?”  All this is there, on the web page, but not there in the web page metadata or, if included, included in the wrong fields and with the wrong punctuation.  And while there may be some logical reason as to why INSTITUTE FOR HISTORICAL REVIEW is all caps in all three references, there seems no logic as to why the APA algorithm failed to find the title of the article;  the MLA and Chicago algorithms did find the article title, even if they generate it as the title of the journal, thus the Italics.

There is a limit to patience and misery.  I have to admit I did not try generating other references. One bad experience is enough; experience says these tools don’t get any better.

What I did do though was to take up the invitation to Check a paper for grammatical errors.  I have a standard paper, an “essay” I compiled some years ago to see what Turnitin would make of it.   I use the term “compiled” with care, since the “essay” was 100% copy-pasted from 15 different online sources. (Turnitin didn’t make a lot of it, finding just one match in the entire 1200 word “paper.”)

I tried uploading this bogus paper into Citation Machine’s Proofreader.

Alas, Citation Machine’s Proofreader baulked; the paper would not upload, the “We’re uploading your paper…” message and the spinning wheel just went on and on.

I had better experience with another Chegg product, EasyBib’s Grammar and spelling checker.  The interface is almost the same as Citation Machine’s (but we won’t dwell on that).  This time my essay did upload so the experience was better, to a point.

A pointless point.

While some of the suggestions may well have been valid, many more were questionable.  I was particularly tickled by this piece of advice:  I was advised that

He gained a degree in construction engineering in London and soon devoted himself to literature

would read better if I replaced “degree” with terms such as “much,” “some,” “much of” or “somewhat.” Errm…

With some trepidation I tried signing up to the plagiarism check. Both Citation Machine and EasyBib offer a free 3-day trial and, as I was in EasyBib, it was EasyBib’s trial plagiarism check that I decided to try.  You have to give credit card details and $9.95 payment is then deducted monthly if not cancelled within the trial period.  Would I be able to cancel? – that was my worry.

That trepidation was not necessary –  I could not complete my registration.  As well as credit card details, the form demands a Zip Code and did not accept a UK post code – and I had changed country of residence to United Kingdom.  So I have no idea if EasyBib’s text-match/plagiarism-detection is any more accurate than Turnitin’s.

The whole notion of automated reference generation, grammar & spelling checking and text-matching makes me feel uneasy.

And as I was writing this post, Nature  published an article by German plagiarism guru Debora Weber-Wulff, Plagiarism detectors are a crutch, and a problem. That term “crutch” is so to the point.  Turnitin and similar services negate learning, they give out misleading messages, they provide poor advice.   Weber-Wulff is most concerned that users of Turnitin (etc)  rely too heavily on the face-value of the output and the reports. They do not investigate or think any more deeply, they do not read more deeply, they stop thinking.

Weber-Wulff’s article focuses on the misuse (through misunderstanding and misinterpretation) of Turnitin by editors and professors, but she does make passing reference to students’ use and misuse of Turnitin too.  Students, often without training and guidance in the interpretation of Turnitin reports and unaware of its shortcomings (including the false positives it finds and the false negatives it does not find), come to see the purpose of academic writing as “Beat Turnitin.”

Those low-level reference generators too negate thinking and learning; their output is taken as gospel.  Users become over-reliant on them – like crutches. They replace learning, users do not bother to exercise so cannot process or progress on their own, without those supports, those crutches.

Learning takes time and effort and concentration, exercise takes time and effort and concentration, practice takes time and effort concentration. And eventually the user can dispense with those crutches.  We can then perhaps begin to think of using the more advanced reference management tools such as EndNote and BibTeX, us using technology, not technology using us.

All in all, it makes me feel uneasy that the respected OWL at Purdue has teamed up with the unsavoury Chegg stable.  However independent the OWL remains, it is tainted by association. User beware.

8 thoughts on “Not such a wise OWL

  1. As a proud Purdue alumna, I also am sorry to read of the OWL entering into any partnership with Chegg.

    Everyone should be aware of where your information might end up when you enter the Chegg.

  2. Great article! Just one little comment. When EasyBib suggested using “some, much, etc.”, it was trying to replace “a degree” not just “degree”, as in liking something to a degree. It wasn’t a good suggestion! But it wasn’t as misplaced as you suggest.

  3. Your article has generated a flurry of activity from my email account to colleagues this morning. I hope that they will find it as useful as I have. It also backs up to what I have been saying to the faculty at my school. If they aren’t listening to me, perhaps they will to you. I went on to read the article from Nature and also forwarded that. I have been concerned for some time with the lack of understanding of the Turnitin reports by both teachers and students. I hope that I can now argue more successfully for the introduction of sessions on how to read a Turnitin report. Thank you.

  4. Thank you, Susan, Nadine, all,

    It surprises me that that Purdue is in such financial straits that it needs to “monetize its free content.” Genuinely free content is a form of goodwill and enhances reputation – and Purdue has good reputation. (Self-promotion granted, but a press release last year begins:
    “Purdue ranks fourth best in the nation for Best Value institutions in the United States, based on a survey released Wednesday (Sept. 5) by The Wall Street Journal and Times Higher Education”
    https://www.purdue.edu/newsroom/releases/2018/Q3/purdue-ranks-no.-4-best-value-school-in-nation-in-wsjtimes-higher-ed-ranking.html)

    This deal is another step towards legitimizing Chegg and its less than savory tactics and strategies – and products. (Great post, Nadine – you’ve gone into this more deeply than I did, I am sorry I did not see this at the time!)

    John

  5. Thank you, John, for your insights once again. I know that at my school most people are blissfully unaware of these issues, and most of all the students. I will be acting on your warnings.

  6. I agree with the contents of this article a great deal, and will add that if one looks below the emboldened agreement text, one will also see Chegg informing the user they can and will scour their computer for data to deliver ads specifically designed for the end user. Privacy concerns be damned, such is the AI age.

    There are several other levels of concern, in that APA and the rest change their formatting requirements often, which are seldom updated in a timely manner in auto formatting programs.

    So all and all, good bye Purdue.

  7. Pingback: Thoughts on the OWL/Chegg partnership – ACRLog

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.