Joe Clark: Accessibility | Design | Writing

Response to Appeal PA07-181

I filed an appeal of GO Transit’s refusal to provide public records. Mediation failed, of course, so the process moved up a notch to the investigation stage, in which all parties have a chance to make “representations.” What follows is my response to those representations, submitted 2007.12.12.

Absence of objection

HDI admits that it has no scope over CNIB’s relationship with GO Transit (“the services of the researcher were retained directly by GO Transit under a purchase order”). All of its comments must be considered irrelevant to the current proceedings (with one exception to be noted later). In any event, even if HDI’s opinions carried any weight in the proceedings, HDI favours disclosure (“it is our opinion that the confidentiality provisions do not apply to these representations and we do not object to [their] being shared with the appellant”).

No party has raised an objection to the release of GO Transit’s internal reports, specifically items 1–5, 7, 8, and 10 as listed in the Notice of Inquiry. Hence those records should be ordered released immediately.

Reporting of numerical results is not a “draft”

GO Transit is keen on making a distinction between “draft” and published reports. While I am seeking any drafts issued along the way, by CNIB’s and GO’s admission, the research in question was carried out in two phases, generating two reports. There is a distinction here between a record that gives numerical results and a later written paper that explains them. Those reporting of numerical research findings is not a “draft.” In the context of measurements, numbers are numbers and are final when they’re committed to paper, assuming accurate transcription. Any such records should be released immediately.

HDI never claimed to have expertise in signage

GO Transit claims that it “has retained the services of CNIB and HDI, highly trained professionals with the required expertise in this field.” CNIB makes the claim, easily proven false (see below), that it has expertise in signage typography, but HDI never makes that claim. All HDI asserts is “we were retained by GO Transit to provide program-management services for a large infrastructure program.” HDI does not have “the required expertise in this field” by its own admission. In fact, HDI admits it has to “retain the services of other specialist researchers, subconsultants or suppliers” from time to time.

HDI writes: “The public disclosure of the work that is the subject of this appeal would reveal how the signage policy is being developed.” GO has not established that a “signage policy is being developed.” The entire process is so disorganized, random, and resistant to rational thought that it certainly does not meet any standard of “policy.” But public disclosure would reveal exactly how signage is being developed, yes (QED). (HDI’s immediately subsequent claim, that release would “potentially allow the appellant to benefit,” is false.)

Parties seem unclear on what a “public document” really is

GO states: “GO Transit did in fact post a notice of participation... for volunteers to participate in the CNIB testing phase. Signage used in this testing phase was for the purpose of research and analysis. The only public document is the GO flyer.” Actually, we are discussing records, not documents. Signage, which carries information, is a record under MFIPPA (“ ‘record’ means any record of information however recorded, whether in printed form, on film, by electronic means or otherwise”). GO Transit signs are records and are subject to the Act.

Parties all but accuse me of intended industrial espionage

I found it interesting that GO Transit and CNIB were somehow able to wrangle their outdated Windows/IE6 browsing platform to read some of the documents on my Web site (and, even more interestingly, on my Flickr account) pertaining to signage and wayfinding. They didn’t do a particularly thorough job. Nor did they put much effort into drawing the right conclusion.

With a business partner, I have done research into typeface legibility for signage applications – for GO Transit, no less (e.g., “GO Transit Type Treatment Report”). Contrary to the insinuations of HDI, which believes that whoever pays for research owns it, the document was nonexclusively licensed to GO Transit for internal distribution and remains our property. (HDI seems to think that American work-for-hire provisions apply in Canada. They don’t.) The demonstration signboards were and are our property and I retain them. (By the same token, CNIB asserts copyright over its own work, going so far as to state that GO merely has “implied licence” to use its work. GO Transit disagrees, claiming “GO Transit will own the final report upon completion.”)

I published the research only because it was clear GO Transit was never going to. Research, particularly publicly-funded research, is meant to be released, not buried.

Next, all my other documents that the respondents somehow managed to copy and paste or otherwise print out consisted of commentary on the state of signage and wayfinding, particularly research in those areas. I note that respondents did a spectacularly poor job of it, being so unadept at the Web (explained by their IE6 usage) that they printed out headline lists of montly archives from my Weblog.

The respondents admit, albeit in ungrammatical sentences, that I conduct and present research on typeface legibility at conferences. What they did not manage to figure out is that I also present critiques of such research. I did so as recently as June 2007 at ATypI Brighton, as is fully documented on my site and elsewhere.

I can assure all parties that I have a smaller client list than they believe (or that their representations make them seem to believe). I certainly have a smaller client base than one of their favoured consultants, Sue Hope, who is presumably earning a good living off the present project. I can provide the written assurance that there are no clients on the horizon, or in the reasonably foreseeable future, or even in a wildly optimistic alternate future, for whom any knowledge I gain from reading publicly-funded research would confer an advantage.

Even if there were any proprietary information in the research (a claim I will shortly dissect), I would have no one to peddle that information to.

In fact, had respondent parties done their homework and actually bothered to read my articles, they would understand that I have an extensive background in journalism and criticism. I am interested in the topic of provably functional typography and in the issue of accessibility for people with disabilities. It seems necessary to explicitly state that I will read the public records in question and write critiques of them – including critiques of purpose, methods, and results, and always within the confines of fair dealing under the Copyright Act.

It so happens that there are very few independent experts who even bother to write critiques of legibility research. (Perhaps the other experts are busy working with clients – and, if I understand the assumptions of the present respondents, are equally busy passing on confidential information from other sources to those clients.) When it comes to public-transit signage in Southern Ontario, I am pretty much the only qualified critic. If I don’t assess the research in question, nobody will.

There is no economic interest at stake. I am not now nor will I be in competition with CNIB, let alone GO or HDI. (I don’t provide direct services to the blind, operate commuter railroads, or run construction projects.)

I should also note that GO Transit has no “competitors”; it has a monopoly on commuter transportation services in its territory. HDI has no “competitors” for its contract with GO Transit. In the signage context, the only competitors CNIB has are, unlike CNIB, actually experienced in and qualified for the tasks they are hired to do. In any event, I’m not one of those competitors.

It is seriously in question whether or not CNIB is acting in accord with its mandate as a charity (registration number 119219459RR0003) while aggressively pursuing signage-development work in competition with the private sector.

Finally, while CNIB claims that, “if the Report and information at issue is disclosed, it could reasonably be expected to... prejudice significantly the competitive position of the CNIB,” such would only be the case if the work were inexpertly done, or if it had been carried out in an effort to prove a preordained point, or were slipshod in some other way. The information, if disclosed, would be prejudicial only if it exposes CNIB as incompetent at its claimed task. Qualified researchers doing good science have nothing to hide.

Publicly-funded research must be available to the public

Having dismissed respondents’ claims that I am somehow in this for the money or am trying to make off, James Bond–style, with their own secret methods and research, we turn now to the issue of compelling public interest, which has several parts.

It’s an established principle that publicly-funded research must be available to the public. It shouldn’t be buried from view (left unpublished), which is exactly what would happen were respondents to prevail. When taxpayers pay for research, they expect accountability: They expect the research to be published. That is merely a restatement of §1 of the Act, which names as one of its “purposes” as providing “a right of access to information under the control of institutions in accordance with the principles that... information should be available to the public.”

Now, let’s examine the specific research in question. I believe a sequence of events largely similar to the following took place.

Now, the foregoing is just my supposition. It could be largely confirmed or denied through release of the records. If I am correct, in the first instance GO Transit and CNIB were guilty of junk science (increasingly a trend whenever the topic is a Tiresias typeface). In this scenario, somebody decided up front that Tiresias is what they wanted to use and hired CNIB to produce a result that supported the decision. While that may not be illegal, it is scandalous if true and compellingly needs to be made public. If false, that too needs to be published. I would call it exactly as I saw it if given the records. I would have no hesitation admitting that my supposition was wrong all along in any detail or as a whole.

Accordingly, yes, there is very much a compelling public interest in issuance of the records. First, I am a member of the public. Through my published critique, the public at large will have access to information about the research practices of a public body. Further, publicly-funded research cannot be “supplied in confidence,” as CNIB asserts. (All that really happened, according to CNIB, was an implication of confidence, which fails to meet the standard.) Research must be published so that other researchers can scrutinize and attempt to reproduce the findings. Research paid for by a public body is a public document.

Possible misdirection of public funds

There is an important whistleblower or accountability-of-public-funds aspect to releasing the research records.

HDI strains the envelope of credulity somewhat when it writes:

It is our opinion that there is no compelling interest in disclosing the records that are the subject of this appeal. The research does not shed light on the functioning of government, impact public opinion or affect political choices.

The Act does not limit “public interest” to those narrow topics. In any event, the first claim is false: The research adamantly does “shed light on the functioning of government,” viz its use of scarce public dollars for potentially questionable research and its use of more of those dollars to bury that research.

HDI continues: “Other than the interest shown by the appellant, there does not appear to be any ‘rousing strong interest or attention’ to the issue other than a general desire to see improvements in the signage provided in Union Station.” Even if such were true – HDI has adduced no evidence – I am a member of the public. If masses of members of the public have the “general desire” that HDI mentions, then they will need to be assured that signage throughout the GO system has genuinely been “improve[d].” Such persons would surely be disturbed if, as I fear, public money had been used to gin up “research” with a preordained conclusion. To be consistent with HDI’s own claims, improvements in signage would need to be unequivocally proven. Concealing the records of research leaves the public with the impression that something embarrassing or wasteful is being covered up.

GO Transit ignored an attempt at mediation

GO Transit (and, for that matter, the Information and Privacy Commissioner) ignored my offer to visit a GO office and take unlimited photographs of the test signboards. As such, GO showed no interest in mediation and indicated a clear determination to go right to the wall to conceal public records encompassed under the Act.

CNIB has no expertise in signage research

“CNIB has significant expertise in the study of low vision,” the Institute writes. GO Transit does no dispute the assertion: “CNIB are experts in the field of visual impairments.”

But the research in question is not a study of low vision or visual impairments. It is an assessment of legibility and usability of signage for all transit users who can see a sign, that is, normally-sighted and low-vision people, possibly with different native languages, but positively excluding those with no vision or no useful vision. (CNIB makes this admission: “[T]his Draft Report is Phase I of the study, involving low-vision individuals. Phase II of the study will involve a broader spectrum of potential users of the new Union Station.”) The research is not a test of low vision or visual impairments.

On the face of it, then, GO made a mistake in hiring CNIB, as CNIB’s professed expertise is not what the task requires.

GO Transit states:

While anyone who works with or has studied graphic design may know the different variations of type and font families used in signage, we submit that the professionals we have engaged have a medical research background in vision and numerous years of field experience which enable them to apply sophisticated and comprehensive methodologies and analyze results for people with variable sight differences.

As such, GO Transit misunderstands its own task. Typefaces are not trivial, interchangeable things in which any idiot off the street who has designed his own Christmas card is an expert. The actual task is one of typeface selection and measurement of signage usability for a broad spectrum of the population, not just low-vision people. As such, while CNIB may be expert at the biology or anatomy of low vision, GO attempts to trivialize the typographic expertise involved. If such expertise is so commonplace, it didn’t need to hire CNIB, which, from all appearances, is carrying out the wrong tests anyway.

Next, the best citations the mighty CNIB can come up with to demonstrate its claimed expertise in typeface legibility are unrelated or simply missing. The paper that CNIB calls “Clear Print Literature Review” actually carries the title “An evidence-based review of the research on typeface legibility for readers with low vision.” It was written by an M.A. and was replete with restatements of basic facts (e.g., what a serif is). The sole topics of interest in the paper were legibility of print (and also the subtopic of legibility of medicine bottles). All other forms of typography, even typography on Web sites, was excluded. The paper had nothing to say about signage legibility. (The allegedly extensive literature searches that underlie the paper did not use any word related to signage as a search term.)

The paper’s primary conclusion was that no conclusions were possible: “[W]e found that the overall body of research on low-vision reading and typeface-legibility characteristics is somewhat inconsistent, with an absence of controlled trials.”

CNIB would ignore its own conclusion in its later document entitled Clear Print Accessibility Guidelines, which recommends the use of Arial and Verdana and monospaced fonts for books and other printed materials. (It recommends all those fonts simultaneously. Also, you’re encouraged to design printed materials with yellow ink on blue.) This is what CNIB’s claimed expertise gets you.

Far from being abundantly concerned with strict research methods, CNIB farms out research to non-doctorates, then ignores the research when it fails to meet CNIB’s preconceptions. (I fear a pattern emerging.) CNIB could not even manage to promote the research and the Clear Print pamphlet correctly on its Web site, initially naming the latter Ckear Print.

Parties in this proceeding somehow beat the odds and managed to wrangle IE6 to print out a few articles from my Web site. Parties never managed to locate the demolition of the aforesaid research and pamphlet that I published on my personal Weblog.

The other asserted paper – “Going Places: Access Needs of Visually-Impaired Travelers in Transportation Terminals: Design Guidelines” – was simply not included in the package of representations. If it were as solid as implied, it would have been offered as evidence. This obscure paper – not available at the Toronto Public Library or the University of Toronto library, or online – must say so little about signage that its inclusion would be an embarrassment. Nor does CNIB even give bibliographic citations of its other research, published or unpublished, to back up its claims of expert knowledge of signage typography.

Those two papers are the full extent of the documentation of CNIB’s claimed expertise in signage typography. In other words, CNIB has no documentation of such expertise. An informed observer would conclude CNIB has no such expertise. There’s only one expert on signage typography in these proceedings.

CNIB at no time asserts it has conducted research on signage, legibility, and typefaces in the built environment. From what I can tell, its contract with GO is CNIB’s maiden voyage in that field. And it shows.

While one would not expect GO Transit or HDI to have even a modicum of typographic knowledge or skill (abundantly clear in the typography of their submissions), CNIB further solidifies its appearance of ignorance by submitting atrociously typeset documents, including a long paper on font legibility that uses 10-point Arial on a 7″ measure (with entire pages of bold italic type, sometimes with underlining). Deeds speak. These are people who cannot make a printed page readable, let alone attractive. They also cannot comply with their own published guidelines for accessible print design.

The CNIB is the organization an uninformed client hires when it doesn’t know enough to find a real expert.

Public safety

I remind all parties that I never claimed there was a “grave” public-safety aspect to the records in question. You will probably not lose life or limb in the hypothetical, but possible, scenario in which HDI, CNIB, and GO Transit decided on a typeface up front and wrote research to back up that choice post-facto.

On the other hand, while GO Transit claims that it, “HDI Consultants, and CNIB have considered all of the safety components in this study,” in fact it is my contention that all parties know so little about typography that the research will recommend use of typefaces that aren’t sufficiently better than the current Helvetica. In the case of an exit sign, statistically distinguishable increases in legibility could save lives. By choosing the wrong fonts due to ignorance, the research would fail to identify typefaces with the highest increase in legibility over Helvetica. That could, but probably would not, have a safety implication.

Parties should not overstate the safety angle. I made sure not to.

CNIB’s “proprietary” methods are, probably unbeknownst to CNIB, industry-standard

CNIB seems unaware that credible scientific research must publish its methods. It claims:

In designing the study and preparing the Draft Report, the CNIB devised a methodology, based on its expertise, to study legibility of signage involving individuals with low vision and distilled the results. The CNIB submits that the design of the study and its methodology are so interwoven in the Draft Report that... disclosure of any of the information... would result in the disclosure of information that is exempt under §17.

In fact, CNIB has no demonstrated “expertise” in signage, legibility, and typefaces in the built environment. As such, CNIB researchers are such greenhorns in the field that all the “methodology” they came up with is surely industry-standard.

How do you test typeface legibility of signage? You decide what fonts to test, you decide on materials and colours, you manufacture some candidate signfaces, you qualify and recruit subjects, you ask their opinions or get them to perform tasks, you collate and publish the results. There: That’s your methodology. There is no other methodology.

CNIB seems so oblivious to the fact that what it is doing is self-evident that it actually admits:

Item 2 shows the types of font sizes that the CNIB chose to study and also reflects the responses that were analyzed in the study.... [A]ll the other records at issue reveal the types of fonts and their sizes, the letters and numbers chosen, and their order.

Of course they do. You are testing fonts and signage.

Disclosure would also reveal the methodology of the Draft Report in respect of the type of lighting used, the distance between the viewer and the signage, as well as details about sign construction and elevation. This information is at the core of the proprietary nature of the Draft Report.

Since CNIB is not claiming to invent a new, commercially available, patentable, or otherwise “proprietary” set of systems for lighting, distance measurement, sign construction, or getting up on a ladder and bolting a sign at a certain height, this objection by CNIB is laughable and shows it misunderstands the scientific method. Font, size, lighting, distance, and sign construction and elevation are what are being tested and aren’t “methods.” They are basic experimental facts that would have to be reported in any scientific work.

Nor is any of this information “technical,” as CNIB claims. They’re just picking some fonts and testing how they work for people. They aren’t doing anything that remotely resembles the design of “methods and measurement to disinfect water,” to use the precedent it cites.

To reiterate a previous point, scientific research must document its methods. CNIB cannot simultaneously claim to be carrying out (“technical”) research while insisting its methods are proprietary and cannot be published. CNIB’s claimed worst-case scenario – “the appellant or a competitor could appropriate the terms and methodology of the study and duplicate it” – is exactly how science works. Research is supposed to be reproducible so it can be verified or disproved.

CNIB is not an “employee” of GO Transit

The Act, at §11(b), protects “employee[s]” of respondent agencies (“A head may refuse to disclose a record that contains... information obtained through research by an employee of an institution if the disclosure could reasonably be expected to deprive the employee of priority of publication”). CNIB, as a corporation, is not an “employee” of GO Transit. There is no direct reporting relationship to anyone at GO (CNIB does not have a boss at GO Transit), there is no regular salary from which statutory deductions are made, there was no hiring or firing process, and CNIB does not spend its workdays on GO Transit property. Outside consultants are not “employees.”

GO Transit flatly admits that CNIB is not an employee, calling it one of “the third parties contracted to this work.” The right of priority of publication does not apply.

GO has procedures in place that frustrate publication

GO Transit admits it has carefully developed a set of procedures that will, in many cases, bury public documents from publication. This admission should be of concern to the Commission, as it implies a wilful pattern of actions meant to illegally defy the requirements of the Act. (Emphasis added below.)

It is common practice in GO Transit that draft documents are not made public as corrections are made to the context during the research and discussion phase....

(The absence of a comma in the original means that, after corrections have been made, GO Transit makes draft documents public. But we know that’s not what they mean.)

If not approved, the study remains deferred and is considered nonpublished. Information remains confidential as it is considered nonconclusive.

GO Transit has a public library.... “Preliminary” and “Draft” reports are not subject to public viewing. Only internal staff have access rights to the drafts. Reports submitted by “third parties” identified as “Confidential” are labelled and catalogued as such to protect the public and research students from viewing....

If Senior Management does not approve, recommendations would not be put into effect and the study would remain deferred and may or may not be brought forward, at a later date [comma in original].

GO Transit has admitted it has elaborate procedures that, when carried out, keep publicly-funded documents out of the public eye. This admission alone argues in favour of disclosure and overrides any other objection.

Given GO’s admission of a system that frustrates disclosure of public documents in ways that may be unlawful, what GO deserves are harsh penalties rather than playing nice. Nonetheless, here’s a mediation offer.

No more than 14 days after receiving a preliminary, draft, or final report or any document in question in this appeal, GO Transit must affirmatively decide to publish or not to publish the document in its library. If published, I receive a copy immediately. If unpublished, I receive a copy as a consequence of this appeal. For documents written by actual GO employees (not by CNIB or HDI), this provision allows GO to enact or waive priority of publication.

In this way, GO Transit may continue to carry out its (possibly illegal) procedures while the provisions of the Act are upheld.

Response to questions from IPC

Posted: 2007.12.15

Homepage: Joe Clark Homepage: Joe Clark Media access (captioning, Web accessibility, etc.) Graphic and industrial design Journalism, articles, book