Court of Appeal for Ontario
Date: October 28, 2025 Docket: COA-24-CV-0770
Judges: van Rensburg, Sossin and Gomery JJ.A.
Between
Oludiran Muyiwa Adewale, Bakri Fawzi Abdulrahman Elymani, Ofokansi Uchenna Gabriel, Moosa Hasnin, Okafor Paternus Ifeanyi, Om Prabha Sudhir Moorti, Okereke Njoku Okereke, Olorunfemi Ayodele Olayinka, Lakshmanan Rajkumar, Sharma Sanjeev and Ezekwem Chidi Victor
Plaintiffs (Appellants)
and
Royal College of Physicians and Surgeons of Canada
Defendant (Respondent)
Counsel
Gavin MacKenzie and Brooke MacKenzie, for the appellants
Michelle Gibbs and Samantha Iantomasi, for the respondent
Heard
May 13, 2025
Appeal
On appeal from the judgment of Justice Jane Ferguson of the Superior Court of Justice, dated August 7, 2024, with reasons reported at 2024 ONSC 3380, and from the costs order, dated August 7, 2024.
By the Court
Introduction
[1] The appellants challenge the dismissal of their action for breach of contract. In that action, they sought a declaration that they each obtained a passing grade of at least 70% on a qualifying exam administered by the Royal College of Physicians and Surgeons of Canada (the "Royal College") in 2020 and were therefore entitled to be certified for independent psychiatric practice. The trial judge concluded that it was "not for [the Superior Court] to step into the shoes of the Royal College and essentially regrade the examination" and that it was "psychometrically justifiable" for the Royal College to adjust the scores of all 2020 certification candidates, including the appellants, by reducing their raw score by 6.5%.
[2] We grant the appeal, set aside the judgment below, and direct a new trial.
Procedural History
[3] These proceedings began as an application and were converted to an action. By the time the matter came to trial, the appellants' claim was that the Royal College had breached its contractual obligation of good faith in the grading of the appellants' exams. The evidence consisted of an agreed statement of facts; affidavits of the parties, with transcripts of their out-of-court cross-examinations; and the evidence of two expert witnesses, including their in-court cross-examinations. The focus of the expert opinion evidence, which was contradictory and complex, was on whether the Royal College's grading of the exams was psychometrically justified.
[4] The trial judge's reasons were insufficient. She set out the positions of the parties with respect to whether the grading of the exams was psychometrically justified and summarized much of the evidence on this issue, relying on extracts from the agreed facts and the parties' written submissions. The trial judge did not, however, conduct the necessary analysis of the expert evidence. She concluded that the Royal College's grading was psychometrically justified and may have considered that this was determinative of the outcome of the proceedings. But she did not analyze the appellants' breach of contract claim to determine whether there was a contract between the Royal College and the appellants, and if so, what the terms of the contract were and whether it was breached.
[5] A new trial is required so the appellants' claim in contract can be properly assessed. Although invited to do so by the parties, it is not appropriate for this court to conduct its own assessment of the evidence and to determine the outcome of these proceedings. Nor can this court conclude on the basis of this record that the appellants' claim is doomed to fail, as the Royal College argues, because the remedy sought by the appellants – a declaration that they passed the exam and that they are therefore entitled to immediate certification for independent practice in psychiatry – is not available. The question of remedy, if any, is best determined in a new trial.
Facts
[6] The Royal College is a national not-for-profit corporation that sets standards for physicians practising in Canada. Among other things, it accredits university programs that train resident physicians for specialty practices and administers exams that residents must pass to be certified as specialists. The provincial colleges that license physicians for specialized practice generally require that applicants obtain a Royal College certification.[1]
[7] Each appellant is an international medical graduate, that is, they obtained a medical degree and completed specialty training in psychiatry in jurisdictions outside of Canada and the United States that are approved by the Royal College. After assessing each appellant's training and practical experience, the Royal College determined that they were eligible to seek certification for independent psychiatric practice in Canada by taking the College's qualifying exam.
[8] Candidates for certification have a three-year window, subject to two possible renewals of one year each, to take and pass the Royal College's qualifying exam before losing their eligibility. In the meantime, they can practice psychiatry (or any other medical specialty for which they are qualified) in Canada within a supervised setting.[2]
[9] Each appellant paid the Royal College $4,415 to take the psychiatry exam offered in 2020. The exam was to consist of two components: a written component, followed by a clinical/oral component a few weeks later.
[10] The 2020 written exam was originally scheduled to take place on March 24, 2020. In the wake of public health measures imposed due to the COVID-19 pandemic, the Royal College cancelled the exam on March 13, 2020.
[11] In April 2020, the Royal College notified the appellants and other candidates that the clinical/oral component of the exam would not be required for their cohort. Instead, candidates had to submit a letter from their respective clinical supervisors attesting that they had the necessary clinical experience, the "knowledge, skills, attitudes and judgment necessary for the independent practice" of psychiatry, and "satisfactory moral and ethical standing". If accepted, this letter would entitle each appellant to take a written exam, eventually set for August 26, 2020. In an April 9, 2020 email to candidates, the Royal College advised that: "If you pass the written exam you will be immediately certified."
[12] On its website, the Royal College stated that the pass score for the written psychiatry exam was 70%. In 2020, in a new entry in the Frequently Asked Questions section, it denied that exams would be graded on a bell curve; instead, a post-exam review would ensure that the exam was "criterion-referenced":
Given the current environment, will the Royal College use a bell curve to grade exam results?
No, the Royal College does not use a bell curve to grade exam results. Our exams are criterion-referenced, which includes an in-depth analysis of all examinations after they are written. This is something we do every year, and 2020 will be no different.
[13] The term "criterion-referenced" was not explained, other than to state that this would include an "in-depth analysis of all examinations after they are written." In another section of its website, however, the Royal College provided this further explanation of the post-exam review process:
Exam data verification
A thorough process of data verification and quality assurance of exam results is conducted during and after the exams.
As part of the quality assurance process for all exams a complete psychometric and performance analysis is conducted. During this process, questions that were overly difficult, non-discriminating, or for which new evidence emerged between the time of question writing and exam administration are reviewed. As a result of this review process, some questions may be deleted from the final exam score for all candidates.
Quality assurance of exam questions
A rigorous exam development and quality review process is followed to ensure the validity and quality of the exam questions.
Following the exam, a psychometric analysis of the questions is conducted and items that do not meet psychometric standards are removed from the exam and the exam is rescored without those items. In addition, any items that have been flagged during the exam administration are reviewed and if these items are deemed to be defective for any reason they are removed from the exam.
[14] There was no other information provided to candidates prior to the August 26, 2020 exam, either on the Royal College's website or in its general communications to all candidates, explaining how the post-exam review process might affect exam scores.
[15] On August 26, 2020, the appellants wrote the exam. Several weeks later, the Royal College told them they had failed it.
[16] The appellants later learned that they had each received an initial score of between 70.45% and 75.76% on the exam before any adjustments were made pursuant to the post-exam quality review process. Three exam questions were removed as a result of the review and the exams were rescored without them. Each appellant had an adjusted score of over 70% as a result.
[17] In addition, however, the Royal College's post-exam review resulted in the "cut score" for certification being set at 76.5% and all examinees' scores being adjusted downward by 6.5%. As a result, the appellants' final grades were between 63.95% and 69.26%.
The Appellants' Claim and the Trial
[18] The appellants commenced this proceeding in January 2021. It was originally framed as an application relying on r. 14.05(3)(h) of the Rules of Civil Procedure, R.R.O. 1990, Reg. 194. The Notice of Application referred to the communications between the Royal College and exam candidates, including the appellants, and sought a declaration that the appellants passed the exam and were entitled to Royal College certification. The proceedings were converted to an action on consent.
[19] In their statement of claim, the appellants pleaded that:
Each appellant paid $4,415 to the Royal College "so that they could write the [Royal College's] August 26, 2020 Exam and have it graded in accordance with the grading criteria advertised or otherwise stated by" the Royal College;
There was an express or implied contract between each appellant and the Royal College;
The Royal College breached the contract by failing to grade the exam in accordance with its own guiding principles and standards for doing so;
The Royal College breached its duty of good faith to perform its contractual duties to grade the exam honestly and reasonably and not capriciously or arbitrarily;
The Royal College's contractual breach caused the appellants to receive a failing score for the exam when they should have received a passing score; and
In addition to being denied certification, to which they were entitled, for independent psychiatric practice in Canada, failing the exam had "significant legal and personal ramifications" for the appellants, their families, and psychiatric patients under the care of some of them.
[20] By way of relief, the appellants sought a declaration that they passed the Royal College's 2020 qualifying exam for independent psychiatric practice and were entitled to certification, and "[s]uch further and other relief as counsel may advise and as this Honourable Court considers just."
[21] A summary trial was conducted. In accordance with a consent order, the trial record consisted of an agreed statement of facts, the affidavits of witnesses and transcripts of their out-of-court cross-examinations, and the affidavits of the parties' respective expert witnesses. The experts, Dr. Sidiq Ali (who provided five affidavits) and Dr. Jack Boulet (who provided three affidavits), were cross-examined in court. The parties provided detailed written submissions both before and after the trial. The trial took three days.
The Parties' Positions at Trial
[22] In their written pretrial submissions, the appellants identified the "central factual issue" in the action as whether the Royal College's 6.5% downward adjustment of their scores on the exam was psychometrically sound or was instead a technical error that lacked psychometric justification. They asserted that "the only triable issue relate[d] to the [Royal College's] performance of its obligations, including whether the [Royal College] provided a norm-referenced Exam (as opposed to criterion-referenced), whether the [Royal College] bell-curved the Exam (contrary to the [Royal College's] website's statements), and whether there was any psychometrically justifiable basis for the [Royal College's] adjustment."
[23] The appellants submitted that the Royal College had contracted with them to score their performance on the exam, and that "[i]n doing so, the [Royal College] was under a common law duty to perform its end of the contract (scoring the Exam) in a manner that was reasonable and not arbitrary" or "fraught with errors that impacted [their] passing scores on the Exam." They claimed that the Royal College's failure to meet its common law duty to score the appellants' performance on the exam in a competent and psychometrically sound manner constituted a breach of contract. The appellants submitted that, in the absence of psychometric justification for the downward adjustment of 6.5%, the Royal College had exercised its discretion in standard setting on the exam "arbitrarily and in breach of its common law duty under the contract."
[24] With respect to the applicable legal principles, the appellants relied on case law on the duty to exercise contractual discretion in good faith and supporting their claim for a binding declaration of right as a remedy for breach of contract.
[25] The Royal College's pretrial submissions asserted that the action had "morphed" into an attack on the soundness of the Royal College's grading system. The respondent identified the issues to be determined at trial as whether the appellants had a cause of action; whether the Royal College's standard-setting processes and decisions were capable of being determined by the court; whether there was a contract, and if so what the terms of the contract were and whether it was breached; and the appropriate relief. With respect to the breach of contract claim, the Royal College asserted that the psychometric validity of its standard-setting process had to be considered together with the totality of its conduct in determining whether it acted in good faith. Finally, the Royal College submitted that the court should not step into its shoes and declare that the appellants passed their exam.
[26] After the trial, the parties each provided 60 pages of written submissions largely devoted to a detailed review and analysis of the expert evidence.
The Parties' Positions on Appeal
[27] The appellants contend that the trial judge's reasons are insufficient because: (i) the reasons do not address the cause of action; (ii) the trial judge made conclusory statements about the inability of the court to intervene in the Royal College's decision-making without referring to any of the case law that was cited to her; (iii) the trial judge provided no reason for preferring one expert over the other; and (iv) the trial judge did not grapple with the core question, which was whether the -6.5% adjustment to the exam scores was arbitrary or psychometrically justified. The appellants also submit that the trial judge failed to decide the issues independently and impartially because she copied and pasted large portions of the agreed facts and the Royal College's submissions into her reasons for judgment.
[28] The appellants contend that the trial judge erred and that this court should set aside the trial judgment. It asks the court to conduct its own review of the record, find that Dr. Ali's evidence should be preferred over that of Dr. Boulet, and grant the relief sought based on its own review of the record.
[29] The Royal College asserts that when the trial judge's reasons are reviewed in the context of the record, they show that she carefully considered the evidence and submissions of the parties, and that she properly accepted the evidence of the Royal College's expert. It contends that, although the trial judge gave no reason for preferring one expert over the other, trial judges are "under no obligation to refer to every piece of evidence, or every expert by name, and [have] the discretion to accept some, none, or all of the evidence presented": Levac v. James, 2023 ONCA 73, 89 C.C.L.T. (4th) 27, at para. 82. In response to the appellants' contention that the trial judge was not impartial and that in effect there was a reasonable apprehension of bias, the Royal College submits that a review of the record demonstrates that she conducted herself in a fair and impartial manner throughout the trial.
[30] The Royal College contends that if there was a reversible error, this court should conduct its own review of the record and dismiss the claim.
Analysis
Insufficiency of Reasons
[31] We agree with the appellants that the trial judge's judgment must be set aside as the reasons are insufficient. The trial judge did not address key issues identified by the parties' submissions, including whether each appellant had a contract with the Royal College, the terms of any contract and whether it was breached. It appears from the issues that she did address that she considered those issues sufficient to dispose of the appellants' claims in the action. That is, the trial judge stated that she was deciding the case on the basis that she was not prepared to step into the place of Canada's national organization responsible for writing and administering national specialist certification examinations and that "[i]n addition, [she did] not find any technical errors in how the examinations were scored".
[32] In Champoux v. Jefremova, 2021 ONCA 92, Hourigan J.A. stated, at paras. 18-19:
The law regarding insufficient reasons is well established. Reasons serve many functions, they: (i) justify and explain the result; (ii) tell the losing party why they lost; (iii) provide public accountability and satisfy the public that justice has been done and is seen to have been done; and (iv), permit effective appellate review.
Functionally speaking, a trial judge's reasons must provide "some insight into how the legal conclusion was reached and what facts were relied on in reaching that conclusion": Champoux, at para. 18, citing Longo v. MacLaren Art Centre, 2014 ONCA 526, 323 O.A.C. 246, at para. 63. In cases where expert evidence is critical to determining an issue, the trial judge is also obliged to explain "in some level of detail" why the evidence of one expert is preferred over another: Champoux, at para. 23.
[33] In reaching the conclusion that the -6.5% adjustment of the appellants' and other candidates' exam scores was derived in a psychometrically sound manner and without technical errors in the Royal College's standard-setting process, the trial judge failed to meaningfully engage with the expert evidence.
[34] The trial judge noted that the appellants relied on the evidence of Dr. Ali to challenge the psychometric soundness of the scoring of the exam, while the Royal College relied on the evidence of Dr. Farhan Bhanji (who held the position of Associate Director of Assessment Strategy at the Royal College at the relevant time) and Dr. Boulet regarding the defensibility of the standard-setting process and the pass/fail decisions made as a result of that process. She stated that she accepted and preferred the evidence of Dr. Bhanji and Dr. Boulet.
[35] The trial judge noted that the appellants' primary issue was whether the 6.5% downward adjustment of all candidates' scores on the exam was psychometrically sound or whether it was instead a technical error lacking in psychometric justification.
[36] The trial judge provided a detailed summary of Dr. Ali's opinion that the downward adjustment should not have occurred.
[37] Dr. Ali explained that the Royal College did not use a "criterion-referenced" methodology (in accordance with what was stated on its website) but had instead adopted a norm-referenced methodology. The trial judge referred to Dr. Ali's evidence that adjustments could be made to a test-taker's score in a criterion-referenced exam, but only where that adjustment is appropriately determined on the basis of prior exams with appropriately similar content – which he described as the established psychometric process of "equating". Instead, the Royal College performed a bell-curve adjustment on the scores, which is a norm-referenced approach that derives the passing standard relative to the scores of an internal normative reference group of examinees rather than anchoring the passing standard in the type and amount of exam content mastery. Dr. Ali's opinion was that the Royal College determined the -6.5% adjustment "without considering… the difficulty of the exam items".
[38] The trial judge's summary of the Royal College's evidence largely consisted of a review of Dr. Bhanji's evidence. It expressed the view that "the standard setting process involves psychometricians and standard-setting judges applying their expert knowledge (in their specialty) of the examination content, looking at information about the performance of the candidates on the MCQ [multiple-choice question] Exam, aided by benchmarks, the performance of previous years, and coming up with a decision based on all that information as to where to place the cut score." The trial judge noted Dr. Bhanji's evidence that the exam was criterion-referenced, albeit informed by normative data.
[39] The trial judge referred only briefly to Dr. Boulet's evidence. At para. 66, she said:
Dr. Boulet testified that the Royal College adhered to relevant standards concerning licensure and certification examination and described that the standards are comprehensive and reflect best practices. There are no MCQ Exams that would meet all of the standards. At a bare minimum, the examination should, inter alia, yield reliable scores (which in his opinion, the 2020 MCQ Exam did). Dr. Boulet testified that the idea that all examinations will meet all standards is untenable. No accredited medical education programs would meet all the standards, but this does not mean these programs are poor.
[40] And at para. 81, after referring to Dr. Ali's opinion that the downward adjustment should not have occurred, on the basis that it was psychometrically arbitrary and without any technical or evidentiary justification, the trial judge stated:
Dr. Boulet concluded that the Royal College's pass/fail decisions on the 2020 MCQ Exam are trustworthy; the 6.5 downward adjustment of the candidates' scores is psychometrically justifiable; and the cut score was determined based on the judgments of subject matter experts who relied on their expertise and knowledge of examination content to derive a point on the score scale that separate those who were competent from those who were not. In short, Dr. Boulet concluded that the Royal College's cut score methods are defensible and further concluded that based on the cut score methods of the Royal College, the plaintiffs each failed the 2020 MCQ Exam because their adjusted scores were below 70%.
[41] The trial judge did not explain how she had resolved the contradictions in the expert evidence. Nor did her summary of the Royal College's evidence directly address the expert evidence in relation to assertions that the Royal College had "bell curved" the grades, failed to conduct a criterion-referenced review, and resorted to "equating". Instead, she simply stated that she accepted and preferred the Royal College's factual assertions, which she agreed were supported by Dr. Boulet. These included that: (i) the Royal College's standard-setting process was criterion-referenced; (ii) there is a psychometric justification for adjusting the scores obtained by the candidates (including the appellants) on the exam; (iii) the -6.5% adjustment of the candidates' scores was derived in a psychometrically sound manner and was derived without technical errors in the Royal College's standard-setting process; and (iv) the adjustment of -6.5% was appropriate, was carried out in relation to the panelists' judgment of the difficulty of the exam, does not constitute bell curving and is psychometrically justifiable.
[42] The trial judge committed the same error that occurred in Champoux. While she set out a summary of some of the expert evidence, she did not meaningfully engage with the evidence and explain why she preferred the evidence of Dr. Boulet over the evidence of Dr. Ali. As in Champoux, there were several issues to be resolved on the basis of the expert evidence.
[43] The trial judge failed to explain why she preferred the evidence of the Royal College's expert, Dr. Boulet, over the evidence of the appellants' expert, Dr. Ali, nor more generally why she "accept[ed] and prefer[red]" the Royal College's factual assertions. This leaves the parties and this court unable to discern why the trial judge endorsed the Royal College's submission that the re-grading of the 2020 exam was psychometrically justified.
[44] The trial judge also failed to engage with the appellants' breach of contract claim, when she stated the conclusion that she would decide the case "on the basis that [she was] not prepared to step into the place of Canada's national organization responsible for writing and administering national specialist certification examinations." While the Royal College argued that this would be the effect of allowing the appellants' claim, the appellants emphasized that the basis of their claim was for breach of contract and that the relief they were seeking – a declaration that they had passed the exam – was available as a remedy for breach of contract.
[45] The trial judge did not address these submissions. Instead, after briefly referring to the fact that several of the appellants had previous failed attempts at the certification exams and that several had since failed subsequent certification attempts, she noted that it was not for the court to "essentially regrade the examination."
[46] A claim for breach of contract was the only cause of action asserted. Despite this, the trial judge did not address whether the parties entered into an enforceable agreement, and if so on what terms, whether the Royal College's duty of good faith performance of the contract required it to grade the exam in a psychometrically sound manner, and whether, even if that did not occur, the Royal College in all the circumstances discharged its duty of good faith performance in the way the exam was conducted and graded. The trial judge was obliged to adjudicate the claim as framed ― that is, as a claim in contract ― even if the parties focused primarily on the competing expert evidence in their submissions.
[47] The appellants also argued that the appeal should be granted on the basis that the trial judge's reasons consisted almost entirely of the parties' submissions. Given the conclusions already reached, it is not necessary to consider this ground of appeal. This should not be taken as an endorsement of the trial judge's approach.
A New Trial Is Required
[48] Section 134(1) of the Courts of Justice Act, R.S.O. 1990, c. C.43 provides that an appeal court may: "(a) make any order or decision that ought to or could have been made by the court or tribunal appealed from; (b) order a new trial; (c) make any other order or decision that is considered just." Section 134(6) provides that "[a] court to which an appeal is taken shall not direct a new trial unless some substantial wrong or miscarriage of justice has occurred."
[49] As stated in Champoux, at para. 19:
When reasons do not fulfill one or more of their functions, it may be necessary for the appellate court to send the matter back to the court below for a new hearing. This is hardly a desirable result, given the wasted costs of the first hearing and the delay in having the case determined on the merits. Yet sometimes, this rather drastic result is necessary to protect the judicial system's integrity. [Citations omitted.]
[50] And, as this court observed more recently: "Where the reasons are deficient in explaining the result to the parties, and the appeal court does not consider itself able to do so, a new trial may be needed. This is a case-specific assessment": Willick v. Willard, 2023 ONCA 792, 492 D.L.R. (4th) 109, at para. 59. In Earl v. McAllister, 2021 ONSC 4050 (Div. Ct.), 156 O.R. (3d) 209, at para. 52, cited with approval in 778938 Ontario Limited v. EllisDon Corporation, 2023 ONCA 182, 479 D.L.R. (4th) 653, at para. 11, the Divisional Court noted:
Section 134(1)(a) of the Courts of Justice Act, empowers an appellate court to make an order or decision that ought to or could have been made by the court or tribunal appeared from. Although, more typically, when an appellate court determines that an error had a bearing on the outcome of a lower court's decision, the matter is remitted to that court for determination, in circumstances where the appellate court has a complete record, the final resolution of the dispute has already been delayed, and there is no special advantage in remitting the matter back to the court in first instance, it will be appropriate for the appellate court to finally determine the issues between the parties. [Citations omitted.]
[51] It is not appropriate for this court to retry the issues in dispute in this action based on the existing record. This was not a long trial. It is not in the interests of justice for this court to make the determinations that are required – the underlying issues for trial are both legally and factually complex and were of necessity canvassed only briefly in this appeal. Essentially, what the parties are seeking is a rehearing on a paper record, with this court stepping into the shoes of a trial court. That is not our function.
[52] An additional relevant factor is the importance of this claim to the appellants. They spent years studying, training, and working with the goal of practicing independently as psychiatrists in Canada. They obtained medical degrees in approved jurisdictions and attestations from their supervisors that they had the knowledge, competence, skills, training and personal qualities to engage in such practice. The only additional hurdle was the Royal College's qualifying exam. Given the impact the Royal College's decision that they had not passed the exam would inevitably have on their lives and careers, the appellants are entitled to a reasoned decision by a trial judge on the expert evidence, as well as the other issues, after full consideration of the parties' written and oral arguments.
Disposition
[53] The appeal is granted. The judgment dismissing the action and the costs order are set aside, and a new trial is directed. As the successful parties on the appeal, as agreed by the parties, the appellants are entitled to costs of the appeal in the inclusive sum of $35,000. The costs of the original trial as well as the costs of the new trial are reserved to the new trial judge.
Released: October 28, 2025
"K. van Rensburg J.A."
"L. Sossin J.A."
"S. Gomery J.A."
Footnotes
[1] There are some exceptions. In Ontario, for example, a physician seeking to practice family medicine may be certified either by the College or by the College of Family Physicians of Canada: Registration, O. Reg. 865/93, s. 3, under the Medicine Act, 1991, S.O. 1991, c. 30. Nothing in this appeal turns on this.
[2] The candidates who wrote the 2020 exam were advised by the Royal College that this examination would not exhaust an attempt at sitting the exam, and that the Royal College would extend all candidates' examination eligibility for one year to 2021. In 2021, candidates were given the opportunity to request a deferral to 2022 without losing a year of eligibility.