In mid-2005, Richard Phelps approached John J. Cannell, who had exposed test score inflation twenty years earlier about writing a joint article about that experience and the developments in testing in the years since. Cannell decided to write his own article, which hoped to have "published in conjunction with may article." In an e-mail to TEG Review editor Bruce Thompson, Phelps was glowing about Cannell:

Cannell is a hero--a courageous man who took on the Mongol Empire all by himself, with no knowledge of battle tactics nor any weapons other than his own common sense.  He fought his battles in the 1980s and then gave up, apparently thinking that he had accomplished nothing.  But, his work had an enormous impact.  Probably over half the validation procedures we slog through when we develop tests where I work we would not be bothering with had it not been for Cannell's reports.  He is the prototypical Third Education Group ideal.  (Well, OK, maybe he and you can be co-proto.)

As the Cannell article progressed, Phelps argued the article should by-pass the review process ("I believe he has the standing to say and write whatever it pleases him to say or write"). Thompson, however, arguing both that the journal was labeled as "peer-reviewed" and that reviewers would be able to adjust their standards to a retrospective essay, sent it to two reviewers: Phelps and Richard Innes.

Both reviewers had some minor suggestions for improvements and recommended publication. Innes also offered to work with Cannell to make sure statements about specific tests were accurate. Phelps was complimentary towards the Innes review: "it looks like a terrific, thoughtful, and helpful review." Thompson asked Cannell to revise his article in view of the reviewers' comments and resubmit it.

In his revision, Cannell added a good deal of additional material, particularly raising questions as to whether present tests suffered from problems similar to those he had discovered twenty years before. Phelps objected strongly to it:

It is wrong.  It is terrible.  It contradicts much of what we stand for.  It completely contradicts my paper.......I would hope that you delete the file, all copies of the file, all references to the file, and all future references to the file.  Wipe its seed from the earth, as it were.  It can do us no good, and it can do us much harm. 

Thompson then asked both the original two reviewers and a third to review the article. Since Innes still strongly supported the revised paper, the third reviewer was needed, because of the now widely divergent views of the first two. Innes and the third reviewer (who remains anonymous but is still listed as a member of Phelps' advisory board) supported publishing the article. Phelps critique was blistering.

Phelps also complained of disparate treatment:

When I submit a paper to the TEG Review, I have to please four reviewers.  When someone else submits a paper to the TEG Review, they need to satisfy only two reviewers, unless you feel you need to dig up more in order to override my vote.

Increasingly he also blamed Innes for the changes that Cannell had made:

I think that you should have informed Richard Innes that Cannell was submitting his essay as an essay, and not as a research paper.  In that case, hopefully, Innes would not have suggested that Cannell dress it up to look like a research paper. 

With two of the three reviewers now agreeing it should be published but the third strongly opposed, Thompson suggested that he publish Cannell's paper, along with a response from Phelps. Phelps initially agreed, "I'm with you on the Cannell-Phelps point-counterpoint. Terrific idea." (Oddly, in light of his opposition to the third reviewer, he also made the following suggestion: "When my firm does open-response item scoring, we have two scorers and if they do not agree on a score, we call in a third, "chief" reader, who gets the deciding vote.")

Thompson then informed Cannell that his paper would be published along with a response. Cannell agreed.

However, Phelps then changed his mind:

Bruce, I've had second thoughts about the Cannell paper. I have a lot to do and I don't want to waste time on this. Moreover, I'm not so sure that I want to be associated with the TEG Review if it is determined to publish stuff like this.

Thompson then published the Cannell paper and (eventually) a response from Doug McRae. Phelps' first response was an announcement: "I am resigning from the TEG Review" and "I want my name taken off the TEG Review editorial board immediately." He then changed the password on the web site.

The next day, Phelps issued an ultimatum, two points of which included:

7) Cannell and Innes will be disinvited from TEG, as will any other reviewers who show a proclivity to sabotage TEG's standards and reputation by pontificating on topics in which they have no expertise.

8) Cannell's paper will not be posted in any form.

Despite his opposition to publishing the Cannell paper, Phelps has since reprinted it on his web site, along with the McRae critique and a critique by him, oddly in light of his earlier snub of this very proposal. In his critique, he continues to blame Innes for the changes in the paper. He accuses Innes of co-writing the article, of being the "ghost co-author." This charge appears to be entirely untrue. Both Cannell and Innes report they had no contact until after the paper was published and did not even have each others' contact information until that time.

In a later e-mail to people who had volunteered for the editorial board expanded on this theme of Innes ghost-writing Cannell's paper:

To provide some context and some contrast, consider the article I submitted to the Third Education Group Review several months ago.  It was reviewed by four reviewers, all of whom accepted it (a couple on the condition of revisions).  I responded to each and every one of their comments and suggestions, along with those of the editor, and all this took quite a lot of time and effort.  But, to any observer of the Third Education Group Review, my article, as well as the earlier one by Sandy Stotsky, despite all the additional rigor of the review process they survived, would appear to be considered of no better quality than the Cannel/Innes article that was rejected by one of two reviewers twice, had a reviewer as a co-author, ignored the comments of a reviewer and the editor, despite ...all the other factors that are listed in the bullets above.  To the casual observer of the Third Education Group Review, there would be no reason to assume that the two articles did not go through review processes of similar rigor.  

Cannell is one of my heroes for what he did twenty years ago.  I searched him out last summer and convinced him to become active again and to write an essay for TEG.  He did and it was fine – his recollections and no more.  Bruce wouldn’t print it – it had to go through review, according to him, even though there was nothing to review—it was just Cannell’s own personal recounting of events, intended for posting as a commentary and not as a research piece.  So, Bruce sent it out for review – to Richard Innes, a man with his own agenda.  Innes started giving Cannell advice and references for making it a “serious” research piece.  The end result is, essentially, an Innes paper, with references to Innes causes that Cannell has never had anything to do with, and wild false assertions and accusations. 

It is very sad.  Innes has used Cannell as a stooge.  And, Bruce Thompson encouraged all of it.

When I first talked to Cannell, many months ago, he emphasized strongly that he did not want to write anything unless it helped my cause.  He even asked me to tell him what to write.  I responded that it wasn’t my place to write for him; he should write his own thoughts.  Apparently, Richard Innes shares no such reluctance.