Tribunal decisions

The decision

IN THE UPPER TRIBUNAL
IMMIGRATION AND ASYLUM
CHAMBER
Case No: UI-2022-002694
UI-2022-002695

First-Tier Tribunal No: HU/06476/2020
HU/06494/2020

THE IMMIGRATION ACTS

Decision & Reasons Issued:
11th March 2024

Before

THE PRESIDENT, MR JUSTICE DOVE
and
UPPER TRIBUNAL JUDGE MANDALIA

Between

SECRETARY OF STATE FOR THE HOME DEPARTMENT
Appellant
and

VARKEY & JOSEPH
(NO ANONYMITY DIRECTION MADE)
Respondent

Heard at Field House on 18 to 21 December 2023

Representation:
For the Appellant: Mr Z Malik KC, Mr C Thomann and Mr R Evans, Counsel, instructed by the Government Legal Department

For the Respondent: Mr G Ó Ceallaigh, and Mr A Schymyck, Counsel, instructed by Imran Khan & Partners Solicitors

Decision and Reasons
Introduction
1. The appellant in the appeal before us is the Secretary of State for the Home Department (“SSHD”) and the respondents to this appeal are Mr Varkey and Ms Joseph. However, for ease of reference, in the course of this decision we now adopt the parties’ status as it previously was before the First-tier Tribunal (“FtT”). We refer to Mr Varkey and Ms Joseph as the appellants, and the Secretary of State (“SSHD”) as the respondent.
2. This is yet another appeal in which this Tribunal has had cause to consider the issues that arise following the filming by BBC Panorama of government-approved exams required to satisfy the English Language requirement in the Immigration Rules. We do not burden this decision by reciting the findings and conclusions reached by successive Presidential panels of the Tribunal. In Ahsan & Others v Secretary of State [2017] EWCA Civ 2009, Underhill LJ referred to the TOEIC litigation to date at paragraphs [23] to [33] of his judgment.
3. We are grateful to Counsel for their clear and helpful submissions, both in writing and at the hearing before us although we have not found it necessary to refer to each and every point they raised.
The Agreed Facts
4. The appellants are husband and wife and are nationals of India. Mr Varkey, prior to leaving India, obtained a BA degree in English Literature and a Diploma in General Nursing, before working as a nurse. He entered the United Kingdom on 27 August 2010 with entry clearance as a Tier 4 (General) Migrant for the purposes of studying an NVQ Level 3 qualification valid until 27 June 2012.
5. Between 21 August 2010 and 28 April 2012 Mr Varkey sat the International English Language Testing System (“IELTS”) secure English language tests. On each occasion, he failed to attain a minimum B2 proficiency score of 5.5 in a single module, reading. However his score in the speaking test was, on every occasion, at or above the required level.
6. Mr Varkey left the United Kingdom on 27 June 2012 and returned to India. He applied to return to the UK and on 10 August 2012, was issued with a Tier 4 student visa which was valid until 27 June 2013. In support of his application for entry clearance, Mr Varkey relied upon a TOEIC test certificate issued by ETS, purporting to verify his sitting and obtaining listening and reading scores of 475 and 410 on 21 May 2012. He claims that he also sat a speaking and writing test on 21 May 2012 that he was later told he had failed, but in respect of which he never received any score. ETS has no record of any speaking and writing test completed by him on that day. In support of his entry clearance application, Mr Varkey also relied upon a TOEIC test certificate which purported to attest to his taking a ‘speaking and writing’ test at the London College of Social Studies on 19 June 2012, and to his attainment of a score of 200 in the speaking component, and 170 in the writing component. Mr Varkey returned to the UK on 25 August 2012.
7. Ms Joseph was granted a dependent visa to join her husband in the UK on 23 October 2012, valid until 27 June 2014. She has since been granted leave to remain in line with that granted to Mr Varkey.
8. On 10th February 2014, the BBC broadcast an edition of “Panorama” on BBC1 in which undercover reporters gained access to several test centres within the United Kingdom. At these test centres secure English language tests were being undertaken by persons subject to immigration control for the purpose of making applications for leave to remain. The BBC investigation revealed, via the use of covert recording devices, significant fraud in the taking of such tests. The investigation has received widespread publicity in the media.
9. In particular, it was revealed by the undercover BBC investigation that oral English tests set by ETS which were being taken remotely via computer, were sat not by the actual candidate but by ‘proxy’ test takers.
10. Following the revelations of the undercover BBC investigation, ETS undertook a review of the validity of test scores awarded by it at various centres in the United Kingdom.
11. ETS informed the SSHD thereafter that test scores for a large number of test-takers had been cancelled. ETS identified ‘proxy test takers’ via the use of computerised voice recognition software, which matched multiple tests taken under different identities. These were then subjected to a further human review in each case by two anti-fraud staff (each of whom has determined that a proxy was used) who purported to determine that the applicant’s ETS language test score was obtained by the use of a proxy test taker.
12. On 17 June 2014, Mr Varkey applied for leave to remain as a Tier 2 migrant, specifically as a registered nurse. This was granted until 19 June 2019. However, on 12 September 2017, his leave to remain was curtailed to expire on 18 November 2017.
13. On 16 November 2017, Mr Varkey applied for further leave as a Tier 2 Migrant. His application was refused on 27 November 2017 on the grounds, inter alia, that he had used deception in his entry clearance application of August 2012. On 3 January 2018 the refusal decision was maintained following an administrative review. The respondent said:
“…ETS has a record of your speaking test. Using voice verification software, ETS is able to detect when a single person is undertaking multiple tests. ETS undertook a check of your test and confirmed to the SSHD that there was significant evidence to conclude that your certificate was fraudulently obtained by the use of a proxy test taker. Your scores from the test taken on 19 June 2012 at London College of Social Studies have now been cancelled by ETS. On the basis of the information provided to her by ETS the SSHD is satisfied that your certificate was fraudulently obtained and that you used deception in your :application of 04 August 20l2.
In fraudulently obtaining a TOEIC certificate in the manner outlined above, you willingly participated in what was clearly an organized and serious attempt, given the complicity of the test centre itself, to defraud the SSHD and others. In doing so, you displayed a flagrant disregard for the public interest, according to which migrants are required to have a certain level of English language ability in order to facilitate social integration and cohesion, as well as to reduce the likelihood of them being a burden on the taxpayer.
Accordingly, I am satisfied that your presence in the UK is not conducive to the public good because your conduct makes it undesirable to allow you to remain in the UK.
In light of this the Secretary of State has deemed that refusal under general grounds is appropriate under paragraph 322(2) and is not prepared to exercise discretion in your favour…”
14. Mr Varkey challenged that decision by a claim for judicial review. Following the grant of permission, that claim was compromised by the parties and the SSHD agreed that Mr Varkey had made a fresh claim under Article 8 ECHR that was likely to attract an in-country right of appeal. The SSHD agreed that, in the event that Mr Varkey succeeds in any appeal on the basis that he did not commit a TOEIC fraud, then, in the absence of a new factor justifying a different course, the SSHD would withdraw the decisions of 27 November 2017 and 3 January 2018 and grant the appellants a reasonable opportunity (not less than 60 days) to submit an application for further leave.
15. In a determination dated 5 April 2022, the FtT allowed the appellants’ appeal against the decision of the SSHD to refuse their applications for leave to remain on family and private life grounds. The decision of the FtT was set aside by Upper Tribunal Judge Canavan and Deputy Upper Tribunal Judge Monson on 3 November 2022. The panel directed that the decision will be remade in the Upper Tribunal.
The Issues
16. The parties have agreed a schedule of issues:
i) Whether Mr Varkey cheated in his TOEIC English language speaking test in 2012, which will require consideration of:
a. The correct approach to the standard of proof where fraud is alleged;
b. Whether the evidence adduced by the SSHD is sufficient to meet the evidential burden;
c. The correct approach that a Tribunal should take to assessing whether, where there is a case to answer, the SSHD’s evidence is to be preferred to that of an individual who denies fraud;
ii) Whether, in light of the answer to Issue i), the appeal should be allowed under Article 8 ECHR.
17. The appellants submit that in addressing issue (i)(b), the Tribunal should address:
i. What evidence in respect of what was actually happening in the college in question, must the SSHD be expected to furnish before an allegation of fraud requiring an answer will be raised;
ii. The extent to which ETS is a reliable organisation that can be simply presumed not to have made an error in producing, transmitting and storing data;
iii. The significance of the fact that a test was taken at a so-called ‘fraud factory’.
18. The appellants also submit that in addressing issue (i)(c), the Tribunal should address:
iv. The relevance of the question of whether a person needs to cheat;
v. The relevance of a delay in the accusation being made and/or a right of appeal being provided.
19. The appellants claim that in relation to these generic issues, it will be necessary for the Tribunal to re-visit the conclusions reached in DK and RK (ETS: SSHD evidence; proof) India [2022] UKUT 00112 (IAC) in light of the additional disclosure provided by the respondent and the additional expert evidence, in particular that of Christopher Stanbury.
Our Approach
20. We do not accept that this Tribunal should address each of the generic issues that we are invited to, as set out in paragraphs [17] and [18] above. Whilst we accept it is helpful for us to address the general principles and the evidence that we have heard and has been drawn to our attention, the evidence relied upon by the parties in a particular appeal is a matter for the parties and is fact specific.
21. As in this appeal, the single factual question that will arise in all appeals such as this is whether or not the appellant cheated in the TOEIC test in question. The appellants accept there is no room for doubt that there was widespread cheating in TOEIC tests. The appellants case is that all cases are not alike and the evidence varies significantly as between colleges and as between individuals.
22. It would be wholly inappropriate for us to provide prescriptive guidance setting out, for example, the evidence that must be provided by the SSHD in respect of what was actually happening in the college in question, before an allegation of fraud requiring an answer will be raised. The evidence will vary from college to college. In some cases, the evidence of a fraud will be perfectly obvious and the individual will be complicit, whereas in others, the fraud may be more discrete and in some cases, as is claimed by the appellants here, unknown to the individual.
23. Providing prescriptive guidance in the way we are invited to by the appellants risks being interpreted as exhaustive or considered to be a bare minimum and is inimical to the interests of justice. As will be apparent from this decision, a fact sensitive analysis of the evidence is required. We therefore focus upon the issues as set out in paragraph [16] above. Our consideration of the issues set out in 1(b) and (c) will however require us to have some regard to the extent to which ETS is a reliable organisation and the significance of the fact that a test was taken at a so-called ‘fraud factory’.
The Appellants case in summary
24. The appellant’s central submission before us is that although there is no room for doubt that there was widespread cheating in TOEIC tests, it is quite wrong to treat the TOEIC scandal as ‘a fraud’. There were in fact numerous different and apparently unrelated frauds carried out in the context of a system in which security was both very lax and entirely dependent on the probity of seemingly unvetted independent contractors. The appellants rely upon evidence that they claim vividly demonstrates the confusion of the small number of ETS staff trying to figure out what was happening at the time.
25. The appellants claim the evidence demonstrates that various frauds were perpetrated in at least three different ways:
(i) direct use of proxies who would ‘swap in’ for those taking the tests who would simply stand aside, as in the Panorama film (the ‘swap in’ method);
(ii) the use of ‘hidden rooms’ in which ‘TeamViewer’ or equivalent software was used to allow a bank of proxies to control the computers of others in an ‘open’ room (the ‘TeamViewer’ method);
(iii) the use of ‘hidden rooms’ in which the test was downloaded onto two sets of computers, but the tests uploaded to ETS came from a bank operated by proxies, rather than those in the ‘open’ room (the ‘parallel test’ method).
26. The appellants claim there appear to be other possible methods, including potentially the recording and reuse of voice files, direct file manipulation, and perhaps more yet to be discovered. There are a range of ‘unknowns’ and different colleges will have adopted different methods at different times that were perfected, or as the case may be, as they became more complacent.
27. Each method has its advantages and disadvantages. In the use of some methods the individual wishing to rely upon a test taken by a proxy will be complicit in the fraud. The two methods that involve the use of a ‘hidden room’, the appellant’s claim, make the detection of fraud difficult, and the appellants accept, would have required some technical skill. The appellants claims the ‘parallel test’ method was much more difficult to detect than the ‘TeamViewer’ method, and had the advantage that it can be operated so that it would not be necessary to turn away honest test takers – they could simply sit the test for the ordinary fee, and if they passed, as they inevitably would, they would not notice anything amiss.
TOEIC Testing
28. Before turning to issues it will assist if we say a little more about TOEIC tests. We have unchallenged evidence before us, which we accept, regarding the form of the TOEIC tests from two sources. The first is set out in a witness statement prepared by Raymond Nicosia, an Executive Director of ETS. The second is the evidence of Richard Shury who was a Security and Compliance Specialist for ETS Global between December 2012 and November 2014. He has provided a witness statement dated 3 June 2023 and gave evidence before us. The evidence establishes there are two separate components to the English language proficiency test:
a. There is a two hour papers based ‘listening’ and ‘reading’ test. The ‘listening’ test was forty-five minutes long and consisted of questions played aloud with answers recorded on a multiple choice answer sheet provided to the candidate. The ‘reading’ section of the test was seventy-five minutes long and consisted of questions answered from a test booklet and the answers were marked on a paper sheet.
b. The ‘speaking’ and ‘writing’ test was computer based, and lasted a total of eighty minutes. The ‘speaking’ section lasted twenty minutes, and candidates wore a headphone with microphone attached. The candidate responded to audio and visuals on the screen, by speaking into the microphone. The ‘writing’ element of the test lasted sixty minutes, and required the candidate to type answers in response to questions on the screen.
29. The evidence before us is that candidates were able to take the ‘speaking’ element only (“a speaking only test”) and that test was run separately and at different times to the ‘speaking and writing’ test. The date and time for ‘speaking and writing’ tests were planned in advance and set by ETS Global. The test would be controlled remotely by ETS Global and delivered to a number of test centres across the UK at the same time. There was one morning and one afternoon test on the first, third, and fourth Wednesday, and the third Tuesday of each month.
30. The time for the ‘speaking and writing test’ was set by ETS Global, as was the time allotted for each question, and could not be altered. The test centre would book with ETS Global, the number of students that were booked to sit the test at the centre. The day before the tests, the Test Centre Administrator’ (“TCA”), an individual trained by ETS Global staff, or the nominated person at the test centre, would have a set of codes sent to them by ETS Global, at an agreed email address. The codes sent were generated for a specific session and were unique to that session.
31. In addition to a computer for each candidate, there was another computer present called the “Admin PC” that was used by the TCA or a ‘Proctor’ to download the test to be taken, and then upload the completed tests.
32. As to events on the day of a ‘speaking and writing test’:
a. An hour before the test, the codes sent by ETS Global would be input by the TCA into the ETS Global website via the Admin PC. This would enable the TCA to download the test material onto the Admin PC, and generate a password referred to as the Proctor Password (a four digit PIN).
b. The individual computers on which the tests were taken would then be turned on, and by visiting the ETS Global website, the TCA could ‘log-on’ to the test and the start screen for the TOEIC test would be displayed. The process could be completed by the TCA or a ‘Proctor’.
c. All the candidate computers are linked to the Admin PC by a ‘Local Area Network’ which negates the need for them to have internet access. The candidates should only have access to the test programme log-in screen on their computer.
d. The candidate enters their name (forename and surname), date of birth and an identification number such as their passport or driving licence number. The information entered is checked by the TCA to ensure it has been entered correctly against the identity document that the candidate is required to have with them. The candidate cannot proceed until the candidate’s ID has been checked and the TCA enters the ‘Proctor’ password.
e. When a candidate logs-on to take the ‘speaking and writing test’, a unique identification number is created. The first four digits are the UK international dialling code. The next six digits are the test serial number and the last six digits are the individual candidate number.
f. The Admin PC shows a list so that the TCA can see that the student computers are logged on.
g. The candidates then navigate a series of questions that are not part of the test but marketing orientated. There is a sound check for the microphone and headphones. The candidate will then reach a screen from which they cannot go any further and a countdown takes place. The TCA has no control of the timing of the test as it is generated remotely by ETS so that the test will not begin until the predetermined start time. Any candidate that has not completed the introduction process by the pre-determined test time, is unable to log on for that test session.
h. The test starts with the ‘speaking test’ that is controlled remotely and lasts twenty minutes. There is then an optional three minute break before the ‘writing’ element of the test begins, which lasts one hour. The candidate is required to remain for the total duration of the test (one hour and twenty minutes).
i. The test completes automatically and a screen appears asking candidates to check their answers. Candidates can check that their answers have been both recorded and typed correctly but no changes can be made to the answers given. Once they are satisfied their answers have been recorded, the candidate clicks finish.
j. Once all candidates have completed the test, the TCA will photograph each candidate, either by use of a webcam on the master computer, a computer set aside for that purpose, or via an iPhone. The photograph element requires the TCA to re-enter the codes previously used to initiate the download for the test. The photograph is married up with the candidate’s details on the computer and the identification relied upon, and then uploaded.
k. Once the ‘speaking and writing test’ is completed the computer based portion is sent to ETS and the test is subject to anonymous marking. The tests were initially transmitted via a web based platform and that was later changed to a mobile platform. The tests were sent to ETS via the YBM platform in South Korea and stored on servers at ETS, backed up by servers in Baltimore.
l. Once the test has been marked, if the test was taken at a college, the score reports were sent to that testing centre to be sent on to the candidate. The unique identification number (see (e) above) appears on the score report generated and test certificate. It also appears on a ‘results audit’ and the ‘audit recording’. The number is unique to the candidate and any associated voice samples provided during that test administration. If a candidate re-takes the test, the candidate will receive a new candidate number for each re-test.
Types of toeic fraud
33. Although Mr Ó Ceallaigh identified three variations of the fraud in his opening (paragraph [25] above), throughout the evidence before us there is reference to various other methods. The task of a Tribunal is that much more difficult because of the wide-ranging terminology used throughout, and interchangeably. To assist in ensuring that we are clear as to the methods of TOEIC fraud that we are asked to consider, the parties prepared an agreed note on the types of TOEIC fraud possible. They provided us with the following suggested descriptions based upon the evidence before us:
(a) Direct substitution: Where a candidate books the test and a proxy physically takes their place at the PC.
(b) Remote control (includes TeamViewer-type software): Where there is a ‘hidden room’ and a proxy in the ‘hidden room’ takes direct control of the PC in the ‘open room’.
(c) Dual monitor: Where there is a ‘hidden room’ and the proxy does the test in the ‘hidden room’, while the candidate in the ‘open room’ has only a second monitor directly connected to the candidate’s PC in front of them, but cannot control anything.
(d) Parallel testing (cloned manager PC): Where the test is run from two manager PCs, in an ‘open room’ full of candidates and a ‘hidden room’ full of proxies, with only the latter tests being uploaded.
(e) Fake identities (with two rooms, one with cancelled tests): Where the test is done by a candidate in a fake identity and by a proxy in the candidate’s identity. The tests in the fake identities are cancelled pre-upload, so only the tests in the candidate’s identity are uploaded.
(f) Mass replacement of voice files: Where voice files are re-used or recordings are made and then replayed during a test.
(g) Possible other methods.
34. The parties accepted in their closing submissions that:
(i) The methods identified at (a), (b), (c), and (e) operate so that the candidate is complicit in the fraud.
(ii) Methods (b), (c), (d) and (e) involve the use of a ‘hidden room’. Method (d) can operate so that the test can be taken by a genuine candidate who is unaware of any fraud, or by a candidate complicit in the fraud.
35. The focus at the hearing before us has been upon the methods that are set out in paragraph [33]. Mr Nicosia refers to the exercise completed following the Panorama programme to review the ‘speaking and writing’ components of the TOEIC tests and identifies a number of categories where irregularities were found:
a. A different voice appears on a repeat test. For example, if the person sits a test in September and another test is received from the same person in October, but the voices on the test are different.
b. There is more than one voice on the test.
c. The same voice features on multiple tests in different identities.
d. The candidate is taking the test but somebody is sitting beside them telling them what to say. This is a ‘script model’ where all the test takes in the same session give the exact same answer.
DK & RK (ETS: SSHD evidence; proof) India [2022] UKUT 00112 IAC
36. The appellant’s claim that it will be necessary for the Tribunal to re-visit the conclusions reached in DK and RK (ETS: SSHD evidence; proof) India [2022] UKUT 00112 (IAC). The Upper Tribunal addressed; (i) the ‘legal and evidential burdens’ and (ii) the state of the evidence produced by the SSHD in ETS/TOEIC cases.
37. The Tribunal was concerned, in particular, with a certificate obtained by DK from a TEOIC Test at the ‘Universal Training Centre (“UTC”) which had been investigated by the BBC Panorama programme broadcast on 10 February 2014, and a TOEIC test certificate relied upon by RK from New London College. In each case the certificate had been cancelled by ETS as ‘invalid’ and in each case the SSHD had refused an application for leave to remain on suitability grounds, alleging that the test certificate relied upon was obtained ‘fraudulently’. Drawing upon the decision of the Court of Appeal in Adeyodin v SSHD [2010] EWCA Civ 773; [2010] Imm AR 704, the Upper Tribunal held that that the reference to “false” in paragraph 322 of the immigration rules means dishonestly false.
38. The Tribunal addressed the burden of proof at paragraphs [47] to [53] of its decision. We agree with the analysis that is set out.
39. The Tribunal confirmed that in in relation to an assertion of dishonesty in proceedings such as these, the standard of proof is the balance of probabilities. It rejected the submission made that in order to discharge the burden of proof, the Secretary of State would need to offer “cogent” evidence, citing the words of Baroness Hale in Re B (Children) [2008] UKHL 35 at [70]:
“Neither the seriousness of the allegation nor the seriousness of the consequences should make any difference to the standard of proof to be applied in determining the facts. The inherent probabilities are simply something to be taken into account, where relevant, in deciding where truth lies.”
40. The Tribunal said:
“58. The evidence as a whole may consist of elements tending to support the proposition with the greatest of certainty, elements tending to support the proposition with much less certainty, elements that are neutral, and elements tending to undermine the proposition. It is their affect as a whole that counts. There is no requirement that any single element have any particular quality other than admissibility, in order to be allowed to contribute to the whole; and there is no requirement that the evidence as a whole have any degree of cogency except such as actually causes the trier of fact to be satisfied as set out above. If the evidence as a whole has that effect the fact is found; if it does not have this effect it is not found; and in neither case is anything contributed to the discourse by separate evaluation of the persuasive force of each item of evidence.”
41. The Tribunal considered the evidence relied upon by the SSHD to support the case that there was widespread fraud and cheating in ETS centres. The Tribunal referred to a summary from the BBC website of the materials exposed by the Panorama programme, and the evidence given at the criminal trials of some involved in the ETS frauds, to set the context that there is overwhelming evidence of fraudulent activity in a number of ETS centres. It recognised that an individual case can never be proved by evidence of generality, albeit whether the burden of proof is discharged in an individual case falls to be determined against the background of the fact that there were many thousands of results obtained fraudulently.
42. The Tribunal went on to consider the individual evidence in each case that comprises of two elements. First, the voice recognition and second, the integrity of ETS’s process for attributing entries to candidates. As to ‘voice recognition’ the Tribunal concluded that there may be a false positive rate of one per cent, or even possibly three per cent, but there is no proper basis for saying that the false positive rate was or would be any higher than that. The Tribunal concluded that the voice recognition process adopted by ETS is clearly and overwhelmingly reliable in pointing to an individual test entry as the product of a repeated voice. The Tribunal said that by “overwhelmingly reliable” it did not mean conclusive, but in general, there is no good reason to doubt the result of the analysis.
43. The Tribunal considered the criticism’s made by Professor Sommer regarding the continuity of records between the test candidate and the test entry at two stages in particular. The first stage concerns the analysis of the sound recordings received by ETS and the steps taken to convert the recording onto a form suitable for voice recognition analysis. To that end, the Tribunal said that it is clear from the other evidence that certain test centres were providing a fraudulent service to fraudulent candidates who paid them for it. There is no reason at all to suppose that they would be other than extremely careful to ensure that the fraudulent entries were indeed credited to the fraudulent candidates.
44. The second stage concerned the test entries while in the control of ETS. The Tribunal said:
“106. .. A suggestion of dissociation of entry from candidate at that point strikes at the heart of ETS’s analytical process. If there had been mix-ups at that point it would mean that as an examining authority ETS was unable to be sure that it was, in general, able to attribute the appropriate test results to candidates. Whatever may be said about the level of supervision (or lack of it) by ETS in this scheme, it does not appear ever to have been said that ETS’s examining process suffers from this defect. Indeed, any suspicion of it would destroy ETS’s reputation globally. According to Peter Millington’s statement, not challenged by the appellants on this point, ETS is the largest private not-for-profit educational testing and assessment organisation in the world, administering 50 million tests annually in 25,000 test centres in 192 countries. It is responsible in the USA for the SAT, a college admissions test, taken by 3 million students a year. It also administers the TOEFL (Test of English as a Foreign Language) test, the most widely respected English-language test in the world, recognised by thousands of colleges, universities and agencies in numerous countries, including the UK, the USA, Australia and Canada. It is clear from its international role and continued viability and dominance that (outside these cases) nobody seems to be suggesting that it cannot be relied upon to attribute test entries to candidates correctly.
107. Again, we would not say that the evidence has to be regarded as determinative. There may be room for error (although none of the experts involved has detected any error, as distinct from showing that there is room for error). What is clear here is that there is every reason to suppose that the evidence is likely to be accurate.”
45. Having considered the wealth of the evidence before the Tribunal and the criticism’s made of that evidence, the Tribunal reached the following general conclusions:
“127. Where the evidence derived from ETS points to a particular test result having been obtained by the input of a person who had undertaken other tests, and if that evidence is uncontradicted by credible evidence, unexplained, and not the subject of any material undermining its effect in the individual case, it is in our judgment amply sufficient to prove that fact on the balance of probabilities.
128. In using the phrase “amply sufficient” we differ from the conclusion of this Tribunal on different evidence, explored in a less detailed way, in SM and Qadir v SSHD. We do not consider that the evidential burden on the respondent in these cases was discharged by only a narrow margin. It is clear beyond a peradventure that the appellants had a case to answer.
128. In these circumstances the real position is that mere assertions of ignorance or honesty by those whose results are identified as obtained by a proxy are very unlikely to prevent the Secretary of State from showing that, on the balance of probabilities, the story shown by the documents is the true one. It will be and remain not merely the probable fact, but the highly probable fact. Any determination of an appeal of this sort must take that into account in assessing whether the respondent has proved the dishonesty on the balance of probabilities.”
46. Mr Ó Ceallaigh submits the decision is of little assistance given the very limited view that the Tribunal was given of what was occurring on the ground in ETS cases. He submits the litigation thus far has been blighted by the lack of adequate disclosure, whereas we now have, for the first time, evidence that ETS knew as long ago as 2013 that the three frauds described by Mr Ó Ceallaigh in his opening, were in use.
47. We turn now to the issues.
The correct approach to the standard of proof where fraud is alleged
48. It is common ground that where the Secretary of State alleges ‘dishonesty’ by the appellant it is for the Secretary of State to prove it, applying the ordinary civil standard of proof, the balance of probabilities. Mr Ó Ceallaigh refers to the decision of the Supreme Court in Jones v Brimingham City Council [2023] UKSC 27; 3 WLR 343, which post-dates the decision of the Upper Tribunal in DK and RK.
49. In Jones, the local authority applied for injunctions to prevent the defendants from engaging in gang-related violence and drug-dealing activity. The Supreme Court held that Article 6(1) of the ECHR does not require the criminal standard of proof to be satisfied in respect of (a) proof that a person has engaged in or has encouraged or assisted gang-related violence or gang-related drug dealing activity within section 34(2) of the Policing and Crime Act 2009 Act or (b) proof that a person has engaged or threatens to engage in anti-social behaviour within section 1(1) of the Anti-social Behaviour, Crime and Policing Act 2014 Act. Lord Lloyd-Jones (with whom Lord Reed, Lord Hodge, Lord Sales, Lord Stephens, Lady Rose and Lord Richards agreed), referred to the retreat from the notion of a flexible civil standard of proof according to the seriousness of the matters to be proved, and references to a “heightened civil standard” that had been traced by Sir Brian Leveson in the Court of Appeal (Jones v Birmingham City Council [2018] 3 WLR 1695). He summarised the developments in paragraph [51] of the judgment:
“I pause at this point to take stock of these developments.
(1) It is now established that there is only one civil standard of proof at common law and that is proof on the balance of probabilities.
(2) Nevertheless, the inherent improbability of an event having occurred will, as a matter of common sense, be a relevant factor when deciding whether it did in fact occur. As a result, proof of an improbable event may require more cogent evidence than might otherwise be required.
(3) However, the seriousness of an allegation, or of the consequences which would follow for a defendant if an allegation is proved, does not necessarily affect the likelihood of its being true. As a result, there cannot be a general rule that the seriousness of an allegation or of the consequences of upholding an allegation justifies a requirement of more cogent evidence where the civil standard is applied…”
50. Nothing said by the Supreme Court in Jones impacts upon what was said by the Upper Tribunal in DK and RK as to the standard of proof. We address the evidence that is now before us below, but as to the standard of proof, the question for the Tribunal is always whether the respondent has discharged the burden of establishing whether it is more probable than not, that the individual acted dishonestly. Lord Hoffman put the matter in context in Secretary of State for the Home Department v Rehman [2001] UKHL 47:
“55. … The civil standard of proof always means more likely than not. The only higher degree of probability required by the law is the criminal standard. But, as Lord Nicholls of Birkenhead explained in In re H (Minors) (Sexual Abuse: Standard of Proof) [1996] AC 563, 586, some things are inherently more likely than others. It would need more cogent evidence to satisfy one that the creature seen walking in Regent's Park was more likely than not to have been a lioness than to be satisfied to the same standard of probability that it was an Alsatian. On this basis, cogent evidence is generally required to satisfy a civil tribunal that a person has been fraudulent or behaved in some other reprehensible manner. But the question is always whether the tribunal thinks it more probable than not."
51. The task of the Tribunal is a simple one, and the paradigm of the fact sensitive decisions made by the specialist Tribunal. As Lord Hoffman said in Re B (Children) (Care Proceedings: Standard of Proof), at [2]:
“If a legal rule requires a fact to be proved (a “fact in issue”), a judge or jury must decide whether or not it happened. There is no room for a finding that it might have happened. The law operates a binary system in which the only values are zero and one. The fact either happened or it did not. If the tribunal is left in doubt, the doubt is resolved by a rule that one party or the other carries the burden of proof. If the party who bears the burden of proof fails to discharge it, a value of zero is returned and the fact is treated as not having happened. If he does discharge it, a value of one is returned and the fact is treated as having happened.”
52. The inherent probability or improbability of an event having occurred will, as a matter of common sense, be a relevant factor when deciding whether it did in fact occur. It forms part of the natural process of reasoning, but should not detract from a consideration of the wide canvas of evidence, before the Tribunal reaches a decision as whether the fact to be proved, here, dishonesty, happened or not.
Whether the evidence adduced by the SSHD is sufficient to meet the evidential burden
53. The difficulty with the use of the phrase “evidential burden” was addressed in DK and RK. In DK and RK, the Tribunal referred, by analogy, to provocation as a defence to a charge of murder. The burden of negating the defence rests upon the prosecution, but before that burden arises, there must be evidence that could raise a prima facie case of provocation. The use of the phrase ‘evidential burden’ is unhelpful in this context because here, the burden of proof rests with the Secretary of State and the issue for the Tribunal is simply whether that burden is discharged.
54. As the Tribunal did in DK and RK, we have considered whether the evidence that is now available to us, including, in particular, evidence from Richard Shury and Christopher Stanbury is such that, taken at its highest, it could not lead to us, or judge’s faced with such appeals, to conclude that the appellant has a case to answer.
Inherent Probability and Improbability
55. Mr Ó Ceallaigh submits that in considering whether an individual has acted dishonestly in this context, the Tribunal must take account of the inherent improbabilities: (i) that none of the many thousands of people accused were innocent (including the many hundreds found by independent judges not to have cheated); (ii) that ETS’ processes, so badly compromised wherever they can be tested, are otherwise infallible; and (iii) that test operators would successfully run a well-advertised and highly in demand test whereby everyone was required to cheat – without being caught. Mr Ó Ceallaigh submits the evidence before the Tribunal must be ‘strong’ before the Tribunal should consider the allegation proved on the balance of probabilities. There is, he submits, no room for papering over the cracks in the Secretary of State’s evidence.
56. There is little to be gained by any detailed excursion by us into the inherent probability or improbability of any particular factor that may be relevant. The propositions advanced by Mr Ó Ceallaigh are so general that they are nothing more than rudimentary. To say it is inherently improbable that none of the many thousands of people accused were innocent, is to vastly understate the reality when it is uncontroversial that there was ‘widespread cheating’. The act of cheating was not in this context, an ‘improbable event’.
57. Mr Ó Ceallaigh refers to the now much publicised litigation concerning the Post Office Horizon software (“the Post Office litigation”) in which it is now apparent that a large number of people faced serious allegations of theft and false accounting in reliance on computer systems. In Alan Bates and Others v Post Office Limited [2019] EWHC 3408 (QB), Fraser J resolved various issues in group litigation between the Post Office and a number of sub-postmasters, arising from alleged defects in an electronic accounting system which had purportedly caused shortfalls for which the Post Office maintained the sub-postmasters were liable. Fraser J found , at [968 – 978], that the Post Office was wrong to contend that the Horizon system was extremely unlikely to be the cause of shortfalls in branches. He accepted, on the evidence, that it was possible for bugs, errors or defects in the system to have the potential to cause apparent or alleged discrepancies or shortfalls in the sub-postmasters' branch accounts or transactions, and that had happened on numerous occasions.
58. We will return to the evidence before us regarding ETS, its systems and its processes but we readily accept that no system or process is infallible. It would be surprising if any credible witness were able to say that any hardware (mechanical or electronic) or any software, even that with advanced security features, is entirely infallible. That is simply a matter of common sense. A system or process that may appear to be impenetrable is always at risk of some failure, malfunction or manipulation. The difficulty with the comparison that Mr Ó Ceallaigh seeks to draw is that unlike in the Post Office litigation, there is no evidence that there were bugs or glitches in the hardware or software used by ETS, or manipulation of results by ETS itself that caused tests to be erroneously attributed candidates. Here, there is a wealth of evidence that colleges in the UK had devised an operation to ensure that candidates were able to succeed in an English language test that they required to support an application to the SSHD, in which the candidate was complicit. Of the possible types of fraud identified by the parties it is striking to see that in at least four of the methods, it is common ground that the candidate is complicit. That is entirely distinct from the position sub- postmasters found themselves in.
59. We accept it is improbable that colleges ran an operation where everyone cheated. A test centre that achieved a 100% pass rate for every test session would immediately arouse suspicion. However, Mr Ó Ceallaigh’s central submission is that there was widespread cheating. One only has to turn to the reference to ‘fraud factories’ in the judgement of Underhill LJ in Ahsan and Others and the general evidence of widespread fraud that was referred to, at [62] to [66] of the decision in DK and RK to see the prevalence of the fraud and the way in which it operated. In R (on the application of Saha and Another) v Secretary of State for the Home Department (Secretary of State’s duty of candour) [2017] UKUT 17, the Tribunal accepted evidence of widespread abuse of testing by the Elizabeth College Test Centre. The centre was not operating genuine tests under genuine test conditions at the material time.
60. At one end of the spectrum, where the SSHD is able to point to reliable and credible evidence that all TOEIC tests undertaken at a college on a particular day or during a particular period, have been classified by ETS as ‘invalid’, there is an inherent probability that the individual engaged in dishonesty, and that establishes a strong case for the individual to answer. In DK and RK the Tribunal referred to the trial at Isleworth Crown Court of some of those involved in the fraudulent activities at New London College, the test centre that had issued a TOEIC certificate relied upon by RK. At the other end of the spectrum there will inevitably be cases where the evidence points to the college having been involved in tests that ETS have classified as “invalid” or “questionable”. That is not however to say that it is inherently improbable that the college ran an operation where everyone cheated or no-one cheated. The judge will take that evidence into account in the overall assessment of the evidence, but it is unlikely to assist a judge to determine the inherent probability and improbability of a fraud being operated at that college.
The General Evidence of Fraud
61. Mr Ó Ceallaigh submits there are several reasons why we should reach a different view of the evidence that is relied upon by the SSHD to that reached by the Upper Tribunal in DK and RK. In summary, there is now evidence from Richard Shury, the Security and Compliance Specialist for ETS Global. He submits ETS knew as long ago as 2013 that there were other methods of fraud apart from ‘direct substitution’ and that evidence was not shared with the Tribunal in DK and RK. On the evidence before us, the Tribunal’s previous faith in ETS as a reputable organisation, he submits, was misplaced. Second, we have expert evidence from Christopher Stanbury regarding the likelihood of the use of a ‘hidden room’. Using the ‘parallel testing’ (cloned manager PC) method, or the mass replacement of voice files, there is a possibility that a genuine candidate would be unaware of the fraud. In DK and RK¸ the Tribunal did not consider the range of methods by which the fraud could be perpetrated without the knowledge of the candidate.
62. The focus in DK and RK, was upon fraud by direct substitution. The focus of the evidence before us has been upon methods involving the use of a ‘hidden room’.
ETS
63. Mr Shury accepted in his oral evidence before us that he has no expertise of working with computers and or system networks, and that when he joined ETS, he was having to train himself, looking at matters in real time, and then looking back and examining processes. The procedures adopted by ETS depended on the honesty of the staff at test centres. Mr Shury’s role involved conducting audits of test centres, mainly unannounced. He reported to Ahmad Bdour who was a Quality and Security Specialist for ETS Global, based in Amsterdam. Mr Shury was told by Mr Bdour that as part of the security process he had been conducting unannounced audits at UK test centres and had encountered problems. For example, he had seen “proxy test takers” (i.e. Direct substitution”) and other procedural violations which should not have been occurring. Mr Bdour had been keeping a log about certain test centres he had suspicions about. A list of centres that required an audit as a matter of priority was compiled. In December 2012 and January 2013 unannounced audits were conducted at six test centres. They included four in London, one in Manchester and one in Leicester. Mr Shury states he witnessed evidence of ‘remote testing’ and ‘proxy test takers’. Mr Shury states audit reports were sent to a number of individuals at ETS including Ahmad Bdour, but in October 2013 he was informed that instructions were given for the audit reports to be sent to two individuals only, Ahmad Bdour and Tom Parker. He was told in November 2013 that:
“..management were sick of hearing about [him] telling them to close centres and [he] was not to bring it up anymore”.
64. Mr Shury states there were a number of test centres, based on audit results and information received from independent TCA’s that had been sent into test centres, that should have been closed. Issues such as identity checks not being completed properly and rooms not being set out correctly had been identified. He would have liked to see stronger action taken against a lot of the test centres, but his recommendation that test centres be closed was overruled, and so other measures were introduced to increase security and compliance, such as the requirement for an independent TCA to run the test session and the use of ‘sealed security bags’ to prevent tampering with test materials. Those measures remained in place until 19 November 2013 when there was an instruction to stop conducting audits and to stop the use of independent TCA’s.
65. During 2013, Mr Shury states Ahmed Bdour and he conducted around 30 security audits. In terms of the deceptions used they were told of: employees giving students the answers, exam sheets filled out by centre employees, the use of proxy test takers and remote testing. Mr Shury states that he does not think that ETS (i.e. the parent company) was aware of the extent of the problems that were occurring at test centres in the UK.
Expert Evidence
66. Christopher Stanbury’s specialist field is Computing, and in particular, Database programming, Web Site development and search engine marketing. His expertise in those areas is not challenged. In his report dated 2 November 2023, in summary, Mr Stanbury concludes that there are a number of ways in which a candidate might falsify their own test results without the assistance of a test centre:
a. The candidate sends a friend or colleague to take the exam in their place. The friend would have to look sufficiently like the candidate to pass the identity check, and this would also be risky if CCTV or photographic evidence were provided.
b. The candidate somehow gains access to the files and amends them before they are returned to ETS
67. Mr Stanbury also identifies a number of ways in which a test centre can manipulate the test results:
a. The simple substitution of candidate’s as shown in the BBC Panorama programme. This is easily prone to detection, but simple to facilitate.
b. Using a single file to replace a candidate’s recordings. This is cheap to do, but can be easily detected by ETS.
c. Replacing a candidate’s recordings individually with recordings made by proxies. This can be problematic because a candidate’s name is not part of the filename and each question would produce a separate voice file.
d. The use of a hidden room with ‘Team Viewer’ software to control the candidate’s PC. This again is easy to facilitate, but also easy to detect.
e. The use of a ‘hidden room’ on a ‘per candidate basis’. This is feasible but not easy to administer. Only candidates paying to cheat would have their answers replaced by proxies. This would require switching for each PC to determine whether the PC used by the proxy or the PC used by the genuine candidate can be used.
f. The use of a ‘hidden room’ on a blanket basis. This is easier to administer after the initial setup. Only the tests completed of the proxy test takers would be uploaded.
68. As to the types of fraud perpetrated, he states:
“6.1.31 With regard to the "types of fraud", much is speculation given that evidence has long since been lost and information from ETS has been disappointingly sparse. The only three types of fraud of which we have good evidence are the direct replacement depicted in the episode of Panorama where pilots replace candidates in the test room, the use of TeamViewer and the "hidden room" as evidenced in the Project Facade reports for Birmingham and Queensway. I have discussed other possible methods elsewhere in this report but in my opinion there is little compelling evidence for them.
6.1.32 Any explanations of what might have happened at test centres have been limited to speculation by the experts involved or by ETS auditing staff (employed to check the test centres were complying with ETS rules). I have examined a statement made by Richard Shury …and by Michael Isaac Kossew … (ETS trainer and auditor), they believed, as I do, that some form of automated cloning system would have been possible. Richard Shury's later statement …confirms that remote control was definitely being carried out. Richard believed that a candidate would be aware "that his computer was being accessed remotely". I agree that this would likely be the case if remote control software were in use but would definitely not be the case if the hidden room was a separate facility and the candidate's results were being ignored. This would have meant that the candidate might have been unaware that their results were being tampered with by the test centre staff…”
69. Mr Stanbury has for some time held a belief that a ‘hidden room’ method was adopted on a blanket basis so that TCA’s were able to substitute test results from a ‘hidden room’ in place of those from the room in which the candidates were sitting. Following a meeting of experts on 27 July 2016, Mr Stanbury and Professor Peter Sommer agreed that it would be possible to run a simultaneous testing session using proxies that could be in a room anywhere in the world, as a variant of the ‘remote control software’ hypothesis, without the need to use remote control software. Mr Stanbury refers to:
a. An audit report relating to Cauldon College on 25 September 2013 that records that there was a ‘suspicion of remote testing’ at that centre. The auditors saw the test centre experience difficulty uploading the tests completed. The audit report states:
“It is my opinion that the technical issue which prevented uploading, reported before by TCAs at various sessions since the policy of sending codes only to ETS TCAs was enacted, is a ruse to allow the Manager PC to be reset, and the tests from a hidden room to be uploaded in place of those from the room in which the ETS TCA sits. I believe the TC are remote accessing the S&W test for the purpose of submitting fraudulent results.”
b. The Project Façade report relating to ‘College of skills and Learning, Birmingham,’ which confirms that during an audit on 16 October 2013 it appeared as if the candidate’s computers were being controlled remotely through the use of ‘Team viewer’ software, suggesting remote testing was taking place. Subsequent analysis of computers seized from the test centre on 2 September 2014 revealed that ‘Team viewer’ software had been installed. Handwritten documents relating to TOEIC exams were discovered during the search and each document listed the test date and candidate details alongside a different name suspected of acting as a ‘pilot’. Other handwritten documents listed candidate and ‘pilot’ names under the title ‘in the secret room’ and ‘out in public’.
70. Mr Stanbury states:
“6.2.12 I believe it is possible a test centre could have, for example, a test room upstairs and a test room downstairs that worked in parallel (note - technically, the second test room could be in a different building or even country). The proxy (pilot) test takers all sat in the hidden (upstairs) room whilst a mix of the genuine and non-genuine (cheating) test takers sat in the downstairs room. In this scenario, the genuine test takers would have thought they were taking the test but, in fact, their answers were potentially being discarded and the answers of the proxy test takers used instead.”
71. In cross-examination Mr Stanbury said that he has seen evidence that supports his view that it is possible that a system could be operated where genuine test takers would have thought they were taking the test. Mr Stanbury referred to an assessment of the security of the TOEIC ‘speaking and writing’ test through the YBM platform completed in October 2013. There were several incidents of ‘remote testing’. Two particular scenarios are identified:
“A. Real test takers were pretending to answer the questions while the real testing was done by proxy test takers in the same building, this was possibly done through sharing the monitor, the technique used was not revealed to us.
B. The admin will start the test in the real test room, shortly after the test is started in a proxy room and the same test takers details are entered, in the real room the test takers are answering the test so for the outside observer the test seems to be running according to procedures, however at the end the proxy tests are uploaded instead.
During the exam the Manager PC in the real room appeared to be offline, and the manager PC for the proxy room was online, both had the same proctor password displayed, no traces of this malpractice on the YBM back office system can be seen. We have tried to mimic the B scenario in our office, we discovered that the manager PC could be ran on more than one PC, we ran it on three PCs using the same code, however each generated a different proctor password which was not the case on scenario B, in our testing we were able to see that there were more than one PC on the YBM back office system and by looking at the Cleanup date field we could see which test manager PC the test was uploaded from.”
72. This, Mr Stanbury claims, demonstrates that it was feasible to have a hidden room, and that, if used in the way described in the report, then it would be an all or nothing approach. That is, the test centre would have to use either all the results from the hidden room, or all the results from the real room. Mr Stanbury believes that it might be technically possible to mix the results but that would require a lot more technical awareness and would be fraught with danger. It is, he claims, entirely plausible that some of the candidates might be completely unaware that their answers are being ignored. Furthermore, auditors visiting the college would be completely unaware that the hidden room was being used. There would be no tell-tale signs of remote testing (as there would be with TeamViewer). The main test room was essentially irrelevant apart from being a good cover for the fraud.
73. Mr Stanbury believes that although it may be possible to find out a candidate's registration number, it is more likely that the administration staff would take the easier option of replacing ALL the candidate's results with those of proxy test takers on the basis that, if desired, all the candidates would pass the test regardless of whether they were paying extra or not. This would be easier and less prone to error than selectively replacing files for those candidates prepared to pay extra to cheat. This would explain why innocent candidates were caught up in the fraud unwittingly.
74. Mr Stanbury also referred to an email sent by Riaz Ashfaq to Richard Shury on 16 July 2023. The email refers to a college in Walthamstow, Queensway College’. The author of that email provided the following information:
“They keep all the students on the 2nd Floor and the helper (they call Pilot) sit in the other floor, which is other side of the college ( can be called 1 and half floor)…
Both the students and the Pilot log in to system simultaneously. The students are instructed to do everything as if they are doing test. At the same time the Pilots do the same test. They have developed a system in such a way that in the server you can just see one log in. When the exam is finished, they upload and submit the exam given by the Pilots.
They have been doing it even in presence of the person sent by ETS. The system they have setup is so well designed that no body will understand it even if the ETS representative upload and submit the exam.”
75. Finally, Mr Stanbury refers to an email sent by Ahmad Bdour to a number of people at ETS Global including Richard Shury concerning the security of the testing platform on 23 January 2014 after concerns were raised bout the possibility of remote access and the use of ‘multiple manager PC’s’. Ahmad Bdour said:
“Multiple Manager PC’s
we also discussed the issue with the ability of tests running on more than one manager PC using the same test code and in the end uploading only one (which is an issue we came across) and which for us leaves no traces of malpractice on the admin platform, Sanghwan was not able to explain this and requested to have more information, I would like to provide them with:
Test center name and test date of when we were sure this has happened, Queensway, 3 test dates before our last audit..
3 test centers we have doubts about for close observation by YBM”
…”
76. In his evidence Mr Stanbury accepts that the ‘parallel testing (cloned manager PC)’ method would require careful planning by the test centre staff, at least some of whom would need to be acting dishonestly. Although he claims it would not be hugely difficult to set up from a technical point of view, he accepted in cross-examination that it would not be difficult for someone who has experience of setting up network computers and technical expertise. However, once set up, it would be easy to manage. He maintained that a recording of the candidate’s test could be substituted with the ‘proxy test taker’s’ test, without the candidate’s knowledge. He states in his report that if the test recordings were uploaded immediately, there would need to be a ‘mirror PC’ being used to answer the questions with the candidate's "Unique ID". The ‘mirror PC’ would be connected to the internet and the genuine candidate’s computer would not. In cross-examination, he accepted that he said that he has a ‘reasonable idea’ of what was going on, but he has not spoken to anyone directly about what was actually going on.
London College of Social Studies (“LCSS”)
77. We have been provided with copies of two audit reports relating to LCSS. It was inspected by Dianke Cisse on 20 June 2012 at 10:25am. The audit report for that visit states there were two testing rooms available with six ‘test takers’ in each room. The overall impression set out on that report is that “The session has been well administered”. There is however evidence before us from Andy Byrne-Smith, an officer in ‘Operation Kraken’ that no ‘speaking and writing’ tests were uploaded by LCSS on 20 June 2012. We have been provided with a copy of an exchange of emails on 20 June 2012 between Saiful Islam, the IT Administrator at LCSS and Tom Parker, the UK Operations Manager at ETS Global. LCSS claimed there had been a ‘short circuit’ during the morning of 20 June 2012 and so the ‘speaking and ‘writing’ tests that morning would have to be repeated. That was the test session referred to as having been ‘well administered’ in the audit report.
78. Mr Shury visited LCSS on 16 January 2013, together with Ahmad Bdour. Their report is more comprehensive and confirms they arrived at 14:18. Mr Shury entered a room with six ‘test takers’ present. Only one had their ID. He was told by the TCA that the ID documents for the other test takers were being copied. During the ‘speaking’ element of the test, only one of the ‘test-takers’ (the one who had his ID) was speaking. The room was left unattended by the TCA on two occasions, and during one of those, Mr Shury observed saw test takers glance at each other’s screens and appear to communicate. During the test, one of the TCA’s came in and asked one of the ‘test takers’ to leave ‘due to difficulty logging on’. Ahmad Bdour entered the room with a copy of the ‘attendance list’ and they proceeded to check the ID of those in the room. Two of the ID’s matched the test-takers and the attendance list. Two others had provided provisional driver licences and were not on the attendance list. One had provided a national ID which was on the list and appeared to match. The two who were not on the attendance list were asked to leave the room by TCA’s. Mr Shury states he saw a TCA flick a switch, and two of the remaining test takers told him they had been disconnected. He states that at that point, they left the test room, and there was an apparent power outage in the building which lasted a few seconds. Ahmad Bdour told the TCA that all tests were cancelled due to irregularities, and the test takers left the building. Mr Shury was told by Ahmad Bdour that he suspected the sessions were being ‘run remotely’. Having left the building, they realised they may need to check the computers further, so they headed back to the test centre. The audit report records:
“We would have arrived back at around 15:25. We entered the second test room, and looked at the computers, asking questions to the TCAs about how the sessions had been run. At this point, the director of the college took us into this office, and admitted to us that he had allowed the sessions to be run remotely; he showed us the set up of how this was done, via a host computer in a back office, and a bank of computers further inside this room.
We then left the test centre.”
79. In his witness statement Mr Shury states the significance of power outages are that the test would not be uploaded. He also confirms that Mr Bdour had told him that he had seen ‘writing appear on computer screens when the test-taker was not typing’. Enquiries were instigated by Mr Bdour with YBM in Korea to establish whether it would be possible for someone to control a computer being used for a speaking and writing test’ so that a ‘proxy test taker’ could take the test on a candidate’s behalf, without having to be physically present at the test centre. He described what he had seen in the following way:
“In a test center they had 2 testing rooms, one was a hidden room in which they had proxy test takers, I will refer to as room (X) and in the other room they had the real test takers which I will refer to as room (1). In room (1) we noticed that the test takers typing was not in sync with what was typed on the monitor, I held a test taker’s hand the typing went on, I then realized that it was the same for all test takers.
After investigations, we were shown Room (X) which had all the computers that were running in the exam, this is where they have the proxy test takers and the exam software running and it appears that in room (1) they only have a projection on the screens of the real test takers of the screens of the test that is running in Room (x).
Through our investigation we understood that they are not using any special software but a “panel” I don’t exactly understand what that means but I saw a lot of switches and cables in Room (x), I believe that they could not have done this from outside the cabled network as they would if they had the chance, so I believe that all computers in room (1) and (X) are on the same network and connected through Ethernet cables.”
80. The response received from YBM was:
“Our development group has reviewed the case based on the limited information provided below, but found no concrete clue about what this center did for cheating. I was advised that there is a clear limitation in controlling the security by technology, unless the center supervisor is on the test‐taker’s side and cooperates each other (sic) for cheating. At the moment, it is very hard for us to know that any software was used at this center for cheating or dual monitors were used to hide proxy test‐taking….”
81. In a further email Ahmad Bdour explained that he had personally attended the test centre and provided the following information:
“The test administrator was indeed part of this setup and coordinating all the activity as they had a hidden room as I explained earlier in which the proxy test takers were seated in, and the real test takers seated in the real test room and were pretending to answer the speaking section and the typing, while actually all they had was remote access to the computers in the hidden room in which the proxy test takers were taking the tests for them, making any inspector think that they are actually answering the questions themselves.

As far as the test center administrator explained to me when he showed me the hidden room, there was no software involved, they only used “panels” I don’t know what that exactly means, but all the computers in the hidden room were hooked up directly to a switch through ethnic cables, I am not sure if you can share monitors through ethnic (sic) cables…”

82. Ahmad Bdour was told that YBM were unaware of ‘panels’ and that they were unable to ‘capture the situation’ from the information provided.
83. The evidence of Mr Shury, which we accept, certainly raises concerns about the willingness of ETS Global, France, and its management to intervene directly and address or bring to an immediate end, the contracts entered into with colleges, and to tackle the frauds being identified. It is clear that those such as Mr Shury and Mr Bdour continued in their efforts to identify those tests centres about which they were particularly concerned between 2012 and 2013, albeit the resources made available were likely to have been inadequate.
84. In DK and RK, the Tribunal noted the absence of evidence from ETS, albeit ETS had had some direct input into some of the witness statements. The Tribunal said, at [109], there was nothing in the material before it in the appeals under consideration to suggest that the lack of evidence directly from ETS ought to raise any suspicion that evidence that might exonerate the appellants was being concealed. The Tribunal evaluated the evidence without hypothesising about what might be the effect of other evidence.
85. The Tribunal said:
“110. … The witness statement of Ahmad Bdour, who worked for ETS and was involved in inspecting testing centres in the UK and elsewhere, does not help the appellants either. It may show that both ETS and the Home Office were or ought to have been aware that fraud was taking place. But it is very far from suggesting that there was any error in the process of transmission of entries, and of course it is not relevant to the question whether frauds were actually committed by particular individuals, except that it contains personal observations of frauds of the sort alleged in the other evidence.”
86. As to ETS itself, the Tribunal said:
“120. … there is no good reason to think of any error in linking the entries examined and classified by ETS with the entries actually submitted on behalf of the candidates to whom they are attributed. The academic evidence was that the “chain of custody” was not absolutely secure and could have been better. That is a world away from saying that it was not in fact wholly reliable. The assertions are made in wholly general terms and any general assertion of that sort is an assertion that ETS has not been reliable in attributing test entries to candidates. But ETS is not concerned only with centres administering TOEIC to those who need a particular result for UK immigration purposes. ETS is, as we have noted above, an organisation with an international portfolio of important and well-regarded tests. We have not heard of any substantiated claim that in respect of any test they have failed in the most basic duty of an examining board, to ensure that the entry examined and attributed to a candidate is the entry made by or on the authority of that candidate. It is in our view inconceivable that ETS would retain the work that they do, if it was thought that their administration was in this respect unreliable.
87. The number of staff employed by ETS Global to oversee the test centres in the UK was, with hindsight, inadequate. Steps were taken to address concerns regarding particular colleges once it became apparent that there were concerns about the reliability of certain test centres. Mr Shury was clear in his evidence that with more resources and permission from senior management, more could have been done. The reluctance to address the concerns within the higher management of ETS Global may, as Mr Ó Ceallaigh submits before us, be because financial considerations such as the need to recover outstanding debts from fraudulent test centres were prioritised, but we are not prepared to speculate upon the commercial risks that ETS Global may have been exposed to, if it had simply pressed ahead with withdrawing from its commercial contracts.
88. Mr Ó Ceallaigh submits the Post Office litigation has revealed that ‘computer evidence’ must be treated with caution and cannot simply be assumed to be accurate. That is undoubtedly true, but the position in not analogous. The fraud here is not established because of some possible glitch, bug, malfunction or manipulation with the hardware and software operated by ETS, but by positive steps taken to manipulate the test results by those charged with administering the tests by ETS, and the candidates taking the tests.
89. The issues in relation to the manner in which ETS Global internally addressed the concerns regarding the frauds does not in our judgment impact upon the integrity of the process that was put in place by ETS following the revelations in the undercover BBC investigation for reviewing and identifying those tests that are ‘invalid’ or ‘questionable’ at various test centres, or the audit trail by which the voice recordings were attributed to candidates. If ETS Global were seeking to bury their heads in the sand and avoid having to face the extent to which fraudulent tests were being undertaken, it would, if anything, under-report the extent of the fraud.
90. We have no doubt that there may have been weaknesses in the security systems and with the benefit of hindsight, additional security features could have been implemented to ensure that there was greater certainty that the correct candidate had taken the test marked. It is the nature of a fraudulent enterprise that the perpetrators of the fraud will endeavour to be one step ahead of those seeking to combat the fraud. The introduction of additional security features is very often no answer because no security feature would be infallible and there is a wealth of evidence that on the whole, test centres and candidates were working in tandem and implicit in the fraud together. A candidate was able to take the test in the knowledge that they would pass and the test centre was able to benefit from the revenue generated by the fraudulent enterprise.
91. As to the evidence of the fraud itself, in DK and RK, the Tribunal said:
“122. … there is no reason to consider that anybody other than the candidates and the test centres in collusion would have wanted to falsify results in this way. We may test this strand by seeing what opportunities there might be for interference with a genuine entry to turn it (as it were) into a fraudulent one. We use that phrase because it is important to be clear that it would be necessary both to add the fraudulent (proxy) entry to the material examined by ETS and to remove the genuine entry. The genuine entry could not be left with its original identity, because that would prevent the addition of another entry having the same co-ordinates; and it could not be exchanged with another genuine entry because that would result in unexpectedly low marks for the victim of that exchange, with consequent appeals, comparisons of the candidate’s voice and that on the test, and clear evidence of the sort of failure of administration that we have just indicated in our discussion of the first strand as completely absent.
123. What would be needed therefore is a process after a candidate’s genuine entry that could substitute an entry consisting of answers given by a proxy tester. As the evidence before us showed, unsurprisingly, this would require considerable technical ability in breaching the security of the test system. There is no evidence that that could even be achieved. If it could be achieved, it is virtually inconceivable that it would be undertaken without any reward for those taking part. Why should anybody go through the test entries, and take great trouble manufacturing better entries and substituting them, if not at the instance of the candidate? In any event there is not the slightest evidence that anybody did or would act in this way.
124. Professor Sommer speculated that a test centre might act in this way in order to manipulate results and so achieve a favourable reputation. That will not do at all. A college might possibly want to make sure that its students did well. If they did, the college could advertise its success in bringing students to the necessary standard and so get more students. But a test centre could not advertise itself in that way: if the tests are administered properly there is no good reason for examinees at one test centre to do better than at another. Even without advertising, the results, if they became known, would risk exciting suspicion (indeed the level of success at certain test centres was one of the features leading to the analysis in these cases). Anyway, a test centre has no particular interest in getting candidates through a test: if they pass it never sees them again, but if they fail, they have to pay to take the test again.
125. There is no perceptible way in which the proxy test entries could have been inserted in the system after the candidates had taken honest tests; and there is no perceptible reason for anybody to insert or substitute them, except at the instance of the candidate. We are left, therefore, with the time of the taking of the test. The material that achieved notoriety in the Panorama investigation and which was used in the criminal trials as well as in earlier episodes of the ETS litigation in these Tribunals shows what happened there. Two observations need to be made. The first is that it is highly unlikely that any candidate present on one of the occasions when proxies were being used was not fully aware of what was going on. The second is that it is if anything even more unlikely that such a system would then attribute proxy entries to anybody who had not taken part in the dishonest scheme, making whatever payment or other arrangement was in place.
92. Mr Stanbury noted the SSHD is reliant on ETS for proof of fraud. He noted that ETS have tested the files using voice recognition software backed up by human checking and have found files where the same voice is attributed to more than one user's test. They also found files attributed to a single user with more than one voice. In his opinion the explanations are:
a. the files have been wrongly attributed to a candidate, or
b. the voice recognition software and human checkers were wrong, or
c. the files do have common voices across different users.
93. He considers the possibility that the files have been wrongly attributed to a candidate to be remote. He accepts ‘voice recognition’ is an area outside his expertise, but has been considered by Professor French previously. He accepts it seems unlikely that the voice recognition software and human checkers were wrong. Similarly he accepts that given the number of times files have common voices across different users makes it likely that there was considerable fraud being committed. Mr Stanbury accepts he has not seen any evidence to suggest how robust, or not, the lookup tool that is relied upon by the SSHD and its related data might be.
94. We are conscious of the evidence of Mr Shury regarding the lack of adequate oversight of test centres at a time when it was becoming increasingly apparent that there was widespread cheating across a number of test centres. The reliability of the ‘Look Up Tool’ is dependent on the reliability of the internal processes adopted by ETS. However, despite the wealth of evidence before us, as the Tribunal said in paragraph [120] of DK and RK, there remains an absence of evidence that in respect of any test identified, ETS has failed in its duty as an examining board, to ensure that the entry examined and attributed to a candidate is the entry made by or on the authority of that candidate.
95. The evidence has evolved with the passage of time. In his report, Mr Stanbury refers to a number of methods by which a candidate’s speaking and writing test can be substituted. Contrary to what is said by Mr Ó Ceallaigh, there was only one fraud. That is, a candidate cheating in a TOEIC test to ensure they achieved a pass. We accept, as Mr Ó Ceallaigh submits, that there was more than one method by which that fraud was perpetrated, and that different methods may have been operated at different test centres. The parties have identified at least six potential methods.
96. As Mr Stanbury himself accepts, as to the "types of fraud", much is speculation. We accept his evidence that there appears to be good evidence of direct replacement as depicted in the episode of Panorama where pilots replace candidates in the test room (Direct substitution), and the use of TeamViewer (Remote control (includes Team-viewer type software).
97. We accept there is evidence of the use of ‘hidden rooms’ at some test centres, but that tells us very little about the actual way in which the fraud was operated by a test centre at any particular time. Mr Stanbury refers to the fact that Richard Shury and Michael Kossew previously believed that some form of automated cloning system would have been possible. He also refers to the Project Façade reports for ‘Queensway College’ and ‘College of skills and learning, Birmingham’
98. The fact that some form of automated cloning system might be possible is not evidence of what was in fact happening. The difficulty with the contemporaneous audit reports and emails that outline the concerns expressed by individuals is that the language used to describe what they saw is vague. Quite understandably, those individuals will not have had the technical expertise to establish precisely how the tests were being conducted, and they used loose and unspecific language, doing their best to describe what they had seen.
99. An audit of Queensway College on 17 September 2013 revealed that a ‘secret room was identified where ‘pilots’ were taking the ‘speaking and ‘writing’ test on behalf of candidates that were located in the examination room. At the College of skills and learning, Birmingham, an audit visit and analysis of computers seized from the test centre revealed that ‘Team viewer’ software had been installed. This is evidence of the less sophisticated types of fraud involving a ‘hidden room’. The evidence of the operation of more sophisticated methods such as ‘parallel testing’ with a cloned PC manager is less clear, and as Mr Stanbury accepts would require at least some level of technical expertise.
100. Mr Stanbury is certain that it is possible that a candidates voice files could have been swapped or falsified with or without their knowledge. We accept it is possible but that is not the question in either this or any other appeal of this nature. It is trite to say that all of the methods identified by Mr Stanbury and others, and perhaps other methods that have not yet been identified, are possible. The fact that something is possible is not to say that is what happened. Beyond saying what is possible on a general level, as Mr Stanbury was bound to acknowledge, it is impossible for him to say in a particular case what method was adopted by a test centre and whether a particular candidate knowingly participated.
101. Mr O’ Ceallaigh accepts in his closing submissions to us that none of the witnesses is able to say with any confidence, what was going on at test centres in 2012. There is simply no way of knowing. The parallel testing method that has long been suspected by Mr Stanbury is simply one possible explanation for the fact that a person who had no reason to cheat, does not have his voice on the voice file. It is, however, he submits, more than a theoretical possibility.
102. There is evidence of some methods having been adopted by test centres and there is hypothesis about other methods. We accept there are, at least in theory, a range of ways in which the fraud and cheating was capable of being perpetrated. We say ‘in theory’ because although the theoretical possibility of any method cannot be ruled out, and there is a wealth of evidence of the less sophisticated methods adopted in which the candidate is complicit in the fraud, the evidence regarding more sophisticated methods is sparse and borders on simple speculation.
103. Take for example a method referred to by Mr Stanbury in his report. He accepts that in theory, a candidate might be able to gain access to the files on their computer and manipulate the results. That would require excellent computer skills and a poor level of security by the test centre. Having set out what would be required on the part of the candidate and test centre, he states:
“I believe this scenario is so unlikely that, although technically feasible, it can be reserved for a movie plot rather than real life.”
104. We accept the evidence of Mr Stanbury that where a test centre uses a ‘hidden room’, it is more likely that the administration staff would take the easier option of replacing all the candidate's results with those of proxy test takers. Adopting that course would be less prone to error.
105. The difficulty with all the candidates tests during a particular session being substituted with the tests completed by proxies, as Mr Stanbury accepted in cross-examination, is that it would mean the test centre has a 100% pass rate for that session, unless one of the proxy’s has been instructed to perform badly. However, he acknowledges that the use of a hidden room on a ‘per candidate basis’ would, although feasible, be difficult to administer.
106. He suggests there is a possibility and perhaps a probability, that on some days all the tests taken at a college are taken by genuine candidates. Again, that is entirely speculation, and in any event, will not assist any appellant in an appeal such as this because where the test is taken by a genuine candidate without the assistance of a proxy, the result, whether it is a pass or a fail, will not be ‘invalid’, because of the absence of the use of a proxy at all.
107. Mr Stanbury’s evidence that the test centre would only upload the tests completed by the proxy’s taking the test in a hidden room without a genuine candidate being aware, so that all the candidates would pass the test regardless of whether they were paying extra or not, is again, nothing more than speculation. In fact, it is contrary to the evidence. Mr Stanbury relies upon the email sent by Riaz Ashfaq to Richard Shury on 16 July 2013, regarding Queensway College that supports his claim that ‘hidden rooms’ are used. However in that email, Mr Ashfaq states that he had taken his nephew to do the test but his nephew was not allowed to take his own test. He provided the names of those responsible for maintaining the system and marketing, and again confirmed that “They do not allow anyone to take exam on their own”. What this evidence demonstrates is that where the use of a hidden room is operated, whatever sophisticated method is devised to avoid detection, the candidate is likely to be aware that they will be unable to complete a genuine test, and a candidate who proceeds to take a test at such a test centre is likely, as a starting point, to be as complicit as if there were simple ‘direct substitution’. It is evidence of the operation of a ‘fraud factory’, where the test centre has no interest in allowing genuine candidates to sit the test themselves.
108. Standing back, we ask ourselves whether the generic evidence relied upon by the SSHD is undermined by the evidence we now have, and is so tenuous, that taken at its highest, a Tribunal could not be satisfied on a balance of probabilities that an individual has dishonestly cheated. The general conclusions reached by the Tribunal in DK and RK are not in our judgment in any way undermined by the evidence of Mr Shury and Mr Stanbury. We are left in no doubt that in general, there was widespread cheating and test centres adopted the less sophisticated methods available of manipulating test results, working in collusion with candidates. It is possible that another method was adopted by a test centre but an appeal is not determined on what is possible. That something is possible is not to say it is probable. The question for a Tribunal is whether it is more likely than not, that the particular appellant they are considering in the case before them cheated.
correct approach that a Tribunal should take to assessing whether, where there is a case to answer, the respondent’s evidence is to be preferred to that of an individual who denies fraud;
109. In Ahsan, Underhill LJ said:
“23. The evidence supplied by the Secretary of State in the substantive TOEIC cases has developed over the course of the litigation. In the earlier cases she sought to rely essentially on (a) generic evidence, given by two Home Office officials, Rebecca Collings and Peter Millington, about the reports received from ETS identifying results as "invalid" or "questionable", and the methodology underlying those reports; and (b) the use of an "ETS Look Up Tool" to marry up those reports with the case of the individual appellant. These cases were not always well-prepared, and in some the look-up tool evidence was not provided at all, or was provided so late that it was not admitted. In more recent cases, however, the Secretary of State has supplemented that evidence by a report from another Home Office official, Adam Sewell, who has analysed the test results from a number of test centres in London. On the basis of his evidence the Home Office case now is that certain centres were "fraud factories" and that all test results from those centres, generally or on certain dates, are bogus. The centres in question include Elizabeth College, which has also been the result of a criminal investigation, under the name Project Façade.
24. The evidence adduced by individual appellants in rebuttal will obviously vary from case to case. At a minimum they can be expected to give evidence that they did indeed attend the centre on the day recorded and took the spoken English test in person. But that may be supplemented by supporting evidence of various kinds: a frequent theme is that it is said to be demonstrable from other evidence that their spoken English was very good and that they thus had no motive to cheat.”
110. In RAM v Secretary of State for the Home Department [2023] EWCA Civ 1323, Underhill LJ (with whom Coulson LJ and Sir Launcelot Henderson agreed) said:
“13. …[The judge] was entitled to, and did, take as his starting-point in paragraph 42 both the fact that the appellant's result was found to be invalid by ETS – that is, on the basis that the voice was that of a proxy – and the fact that [New London College] was an established fraud factory. Those findings, as the UT makes clear in DK and RK, make it not only probable but highly probable that he had in fact cheated. He was nevertheless obliged to consider the appellant's evidence to the contrary. He recognised that and considered it fully at paragraph 43, focusing on the main point which had been made before him by Mr Bellara related to the standard of the appellant's English. But it is one thing to say that the appellant's evidence had to be considered; it is another to say that it was obliged to be accepted. It was the judge's task to decide whether that evidence outweighed the effect of the generic evidence; and the message of DK and RK is that a mere denial is very unlikely to do so.
14. Mr Bellara says that the appellant's evidence amounted to more than a mere denial, but except in the most literal sense I cannot accept that. It is true that he adds a few details to his account of having taken the test himself - about how he chose NLC as the college to take his test, how he travelled there, and the numbers of people present and the like. He also (a point to which I will return) gives evidence about his proficiency in English. But none of that very limited amplification of the bare assertion that he took the test was capable of casting serious doubt on the reliability of the results. The fact that the oral evidence was given consistently and with apparent conviction, and thus was credible if viewed in isolation, is not enough. The question for the judge was whether it was sufficient to discharge the evidential burden of proof created by the generic evidence, and he was fully entitled to reach the conclusion that it did not.
15. As regards the evidence about the appellant's standard of English at the time that he took the test, which is the subject of ground 2, I can see nothing wrong with the judge's reasoning in paragraph 43 and, certainly nothing amounting to an error of law with which the court could interfere. Mr Bellara effectively acknowledged that this was very much a secondary point in his grounds of appeal, as Snowden LJ had already observed.”
111. To put what was said by Underhill LJ in paragraph [15] in context, the Upper Tribunal judge had considered the appellant’s claim that he had studied the English language in India. The judge accepted there may have been some improvement in his English language ability during the ten months in this country, but went on to note that there may be many reasons as to why somebody with a reasonable command of the English language might use a proxy taker, for example fear of the adverse impact of failure, or a concern as to failure consequent to nerves. See MA (ETS – TOEIC testing) [2016] UKUT 450.
112. Mr O’ Ceallaigh submits the decision of the Upper Tribunal in MA was specific to the facts where MA was hopelessly lacking in credibility on a range of issues. He submits that where a person is entirely lacking in credibility, that cannot be displaced by the fact that they speak good English, but that was elevated in DK and RK, to a point of principle. We do not accept that submission. At paragraph [108] of the decision in DK and RK, the Tribunal was addressing the evidence of Professor Sommer to the APPG that one of the features of evidence that one would look for is corroboration. It is obvious that if an individual displays a grasp of the English language that is lower than that required for the test, that is likely to weigh against the individual. The Tribunal used the phrase “incompetence in English”, and a judge must be careful not to fall into the trap of conducting some assessment of the individual’s competence in the English language. That does not in any way undermine what was said in MA:
“57. … In the abstract, of course, there is a range of reasons why persons proficient in English may engage in TOEIC fraud. These include, inexhaustively, lack of confidence, fear of failure, lack of time and commitment and contempt for the immigration system. These reasons could conceivably overlap in individual cases and there is scope for other explanations for deceitful conduct in this sphere….”
113. A judge has to start somewhere. The evidence before us simply serves to re-enforce the fact that there was widespread fraud and cheating at ETS test centres. As the Tribunal said in DK and RK, it is clear beyond peradventure that where there is evidence from ETS that points to the test relied upon by the individual as having been taken by someone other than that person, that is strong evidence that will weigh against the individual and calls for a credible explanation.
114. The parties agree that a Tribunal must consider the evidence before it as a whole and the decision will be fact sensitive. In reaching its decision, the Tribunal must survey the wide canvas of evidence before it. The factual determination must be reached on the basis of all available materials, and the Tribunal must consider each piece of evidence in the context of all the other evidence.
115. We accept that in many cases there is likely to have been a passage of time, sometimes significant, before the allegation of dishonesty is made by the SSHD. There are important limits as to the reliability of human memory, and a greater confidence displayed by a witness does not necessarily correlate with a more accurate recollection of events. Contemporaneous evidence to support an account remains important and the absence of documents that should be capable of being produced to support an individual account, is a factor that is capable of weighing against the individual. If the account given by witnesses cannot be tested by reference to other evidence before the Tribunal, the judge is likely to place greater reliance upon their assessment of the credibility of the witness overall, and the inherent plausibility (or implausibility) of their account.
116. Having considered the explanation provided by the individual it is for the Tribunal to determine whether the respondent has established that it is more likely than not, that the appellant acted dishonestly.
These Appeals
117. We have already set out the key factual background. There is evidence before us that Mr Varkey obtained a BA degree in English Literature (3rd Division) from the Mohanlal Sukhadia University in Udaipur, India in 2005. He was also awarded a Diploma in General Nursing, by the Directorate of Medical Education in Karnataka, Bangalore, in August 2007. He entered the United Kingdom on 27 August 2010 with entry clearance as a Tier 4 (General) Migrant for the purposes of studying an NVQ Level 3 qualification valid until 27 June 2012.
118. Between 21 August 2010 and 28 April 2012 Mr Varkey sat the International English Language Testing System (“IELTS”) secure English language tests. His scores were as follows:
Test Date
Listening
Reading
Writing
Speaking
Overall
21.08.10
5.5
4.5
5.5
6.0
5.5
28.04.12
5.5
5.0
6.5
7.0
6.0

On each occasion, he failed to attain a minimum B2 proficiency score of 5.5 in a single module, reading. However his score in the speaking test was, on every occasion, at or above the required level.
119. Mr Varkey claims that he had been offered a place to study for an MBA, subject to him meeting the relevant English language requirement. His evidence is that there were no IELTS test slots available either in the UK or in India, and so after some on-line research he became aware of ETS TOEIC tests being offered by the London College of Social Studies. He called the test centre and then attended the centre to book the test. He was given a slot to take the tests on 21 May 2012. He claims that on that day he travelled to the test centre from his home in Seven Sisters. He claims there were around 12 to 15 candidates in the examination room and each individual completed the tests at a computer and were given sheets to mark answers. He claims the centre was very noisy and everyone was speaking loudly which meant he could not concentrate. Mr Varkey claims that he received the score report relating to the ‘listening and reading’ tests about a week later, but no report concerning the ‘speaking and writing’ tests. He telephoned the test centre and was informed that he had not passed the ‘speaking and writing’ test. He asked for the test scores and when he returned to the test centre he was informed that his scores were much lower than required to pass and, that he would have to retake the test. He claims he booked to retake the test on 19 June 2012 and paid for that test.
120. Mr Varkey left the United Kingdom on 27 June 2012 and returned to India. His friend called him in India and he learnt that he had passed the speaking and writing test. He applied to return to the UK. In support of his application for entry clearance, Mr Varkey relied upon two TOEIC test certificates issued by ETS, which set out the following scores.
Test Date
Listening
Reading
Writing
Speaking
21.05.12
475/495
410/495

19.06.12

170/200
200/200

121. We have been provided with a copy of the official score reports relating to both tests. The official score report relating to the appellant’s ‘speaking and writing test’ issued by TOEIC confirms the appellant’s name and date of birth. It records a 16 digit ‘registration number’ relating to a test completed on 19 June 2012.
122. In a witness statement of Darren Morley dated 17 May 2020, which we accept, it is said that the appellant’s test result had been cancelled by ETS on the basis that its own analysis indicated that the test result had been obtained via the use of a proxy tester. Mr Morley states the Home Office was notified by way of an entry on a spreadsheet, an excerpt from which he exhibits. The excerpt refers to the relevant ‘certificate number’ and claims the test is ‘invalid’. Although that excerpt refers to a ‘certificate number’, the TOEIC score reports do not have a ‘certificate number’. The ‘speaking and writing score report’ has a ‘Registration number’ which matches the excerpt from the spreadsheet that is exhibited to the statement of Darren Morley. We are satisfied the test now referred to as ‘invalid’ by ETS is the test Mr Varkey claims that he took on 19 June 2012. Mr Varkey has obtained a copy of the voice recording and he accepts the voice on that recording is not his.
123. Adam Sewell has exhibited to his witness statement, a spreadsheet providing a summary of the TOEIC test data relating to LCSS which includes the test sessions attended by, and attributed to the appellant. Between 15 November 2011 and 15 January 2013 the LCSS completed a total of 1,281 ‘speaking tests’ that were later subject to voice analysis by ETS. Of these, 895 tests were deemed invalid, and 386 were deemed questionable. There were no tests that were found to be genuine. The ‘speaking tests’ were completed during the course of 100 test sessions. Proxy test taking was identified in 97 of the 100 test sessions. The three test sessions in which proxy test takers were not identified were sessions during which there were one to three candidates.
124. On 19 June 2012, the LCSS reported to ETS that the first appellant had sat a ‘speaking and writing’ test. On that day, the LCSS ran a speaking and writing test during the morning that was attended by ten candidates. It also ran a ’speaking only’ test in the morning that was attended by three candidates and another in the afternoon that was attended by six candidates. Mr Sewell states that the voice recordings of all ten candidates that completed the speaking and writing’ test were assessed by ETS and nine recordings, including that of the first appellant, were matched to the voice recordings of other TOEIC candidates. The nine candidates deemed to have used proxy test takers were reported to the Home Office as having ‘Invalid’ results, while the other one candidate was reported to the Home Office as having a ‘questionable’ result.
125. Mr Sewell referred to the claim made by the appellant he had taken a ‘speaking and writing test’ on or around 21 May 2012. He has been unable to find any record of the LSCC operating any ‘speaking and writing’ test on 21 May 2012. The nearest recorded sessions for ‘speaking and writing’ tests at the LSCC were on 16 May 2012 and 23 May 2012.
126. Mr Varkey maintains he completed the tests himself and has a very good English language speaking ability. He refers to the IELTS tests that he previously passed. He also refers to the IELTS test he has since completed, that reveal the following scores:
Test Date
Listening
Reading
Writing
Speaking
Overall
06.04.13
5.5
5.0
5.5
7.0
6.0
26.10.13
5.5
5.5
5.5
7.0
6.0
29.01.15
6.0
5.5
6.0
6.5
6.0

127. In cross-examination, Mr Varkey said his first language is Malayalam, and that when he studied his BA in English Literature, students had a choice of taking exams in either English or Hindi. He chose English. He was referred to the ‘Statement of Mark’ issued by the University. He accepted that in the first year, for English Literature he attained 80 marks. In the second year, he attained 79 marks and in the third year he attained 72 marks. The minimum mark required each year was 72 and the maximum mark possible was 200.
128. Mr Varkey said that he was planning to start an MBA in September 2012 and that is why he sat the IELTS test on 28 April 2012. He said that the test is hard and accepted that he was only achieving modest scores to the lower end of what was required. He accepted that his leave to remain was about to expire on 27 June 2012 and that he only had limited time to successfully complete the test. He claimed that no IELTS slots were available and that is why he chose to complete the ETS test. He accepted he could have completed an IELTS test in India, but claimed he wanted to complete the test as early as possible. He had spoken to the University and he had been advised to search on-line for test centres. He denied that he had chosen the test centre because he was tipped off that there were centres that could guarantee him the required score. He said that when he attended the test centre to book the test, he had paid for the test in cash.
129. Mr Varkey claimed that on 21 May 2012 he had completed the ‘speaking and writing’ test in the morning, and completed the ‘reading and listening’ test that afternoon. He said that when he attended to retake the test on 19 June 2012 he did not notice anything strange. In cross-examination he claimed there were about 12 – 15 people there. When asked about the absence of any information in his witness statement about the test he claims to have taken on 19 June 2012, he denied that that was because he had not taken the tests on that day.
130. We have received witness statements and heard oral evidence from Mr Rodney Ellington, the second appellant, and Mr Jijo Johnny. We accept that each of those witnesses were doing their best to assist the Tribunal, but we gain little assistance from their evidence. There can be no way of knowing the yardstick by which the witnesses called measure the appellant’s English language ability. The question for us is not whether the appellant was competent in his use of the English language in 2012, but whether he acted dishonestly. An ability to speak English does not mean that the individual was bound to pass any particular component of the tests and in any event, as has been recognised previously, there can be a myriad of reasons why even someone that has a reasonable grasp of the English language might nevertheless cheat in a test. We have no doubt that the second appellant sincerely believes that her husband is entirely honest and that he would not cheat, but as she was bound to accept, she was not in the UK at the material time and only joined her husband in November 2012.
131. Mr Stanbury expresses the opinion that the first appellant’s test recording was never made because the LCSS was falsifying all tests, or if it was made, it was lost, replaced or confused with another recording later in the process, either deliberately or accidentally. In either case, that was done potentially without the knowledge or involvement of the first appellant. We are unable to attach any weight to his opinion. Although we accept the opinions expressed by Mr Stanbury regarding matters that are within his expertise, the difficulty with much of the evidence of Mr Stanbury is that he is prone to speculation. His opinion is based upon what he considers to be possible. He accepts however that he did not know what was actually happening at test centres in 2012. Any opinion expressed by him as to what the LCSS did strays beyond his knowledge or expertise. It is not for him to speculate as to what may have happened to any recording made.
132. We reject the first appellant’s claim that he had taken a ‘speaking and writing test’ on the morning of 21 May 2012. That was the third Monday in May 2012. The unchallenged evidence of Mr Shury is that the ‘speaking and writing’ tests occurred on the first, third and fourth Wednesday and the third Tuesday each month. Mr Varkey could not therefore have taken a ‘speaking and writing’ test on Monday 21 May 2012. The spreadsheet exhibited to the witness statement of Adam Sewell providing a summary of the TOEIC test data relating to LCSS confirms that on 21 May 2012, the LCSS completed 18 ‘listening and reading’ tests. His claim in his witness statement that there were 12 to 15 candidates and that it was very noisy is simply not credible. Quite apart from the fact that the LCSS would not have had any access to a ‘speaking and writing’ test on that day because no session had been assigned to them, there is no reason let alone credible reason why the LCSS would have mimicked a ‘speaking and writing’ test that the first appellant, and according to him 11 to 14 other candidates would think they were taking. That significantly undermines his credibility.
133. On 19 June 2012, the LCSS undertook speaking and writing tests during the morning and afternoon. Mr Varkey’s written evidence of that test is devoid of any detail and his evidence overall is vague at best. He said in his oral evidence that there were about twelve to fifteen people there. That, we note, is precisely the same number of candidates that he claimed he had taken the test with previously on 21 May 2012. In fact, his test was completed amongst a cohort of ten test takers. The evidence of Mr Sewell, which we accept, is that the voice recordings of all ten candidates that completed the speaking and writing’ test that morning were assessed by ETS and nine recordings, including that of the first appellant, were matched to the voice recordings of other TOEIC candidates. The nine candidates deemed to have used proxy test takers were reported to the Home Office as having ‘Invalid’ results.
134. Having considered all the evidence before us, we find that having failed in the IELTS test on 28 April 2012 to attain a minimum B2 proficiency score of 5.5 in reading, notwithstanding the score achieved for speaking and writing, the appellant was anxious to ensure that he completed the English language tests to ensure that he would be able to return to the UK after his leave to remain expired, so that he could enrol on an MBA course. He was, we find, aware that he could ensure that he would pass the tests by taking the test at a centre like the LCSS which was involved in widespread fraud and would facilitate a test by a proxy.
135. There is no credible evidence before us that in May 2012, the LCSS either had the expertise or was operating a method of fraud that involved parallel testing with a cloned manager PC. The audit report of the visit to LCSS by Mr Shury and Mr Bdour on 16 January 2013, and the subsequent enquiries that were instigated by Mr Bdour with YBM in Korea, refer to ‘proxy test takers’ taking the test on a candidate’s behalf, without having to be physically present at the test centre. Importantly, what Mr Bdour described to YBM was that they noticed that the test takers typing was not in sync with what was typed on the monitor. Mr Bdour said that he held a test takers hand, the typing went on, and he then realized that it was the same for all test takers. That is not indicative of the sophisticated setup required for parallel testing with a cloned computer, but evidence of the ‘remote control (including TeamViewer-type software)’ method identified by the parties. The parties accept that method operates so that the candidate is complicit in the fraud.
136. Standing back and having considered the evidence before us as a whole, including the evidence of the appellant and his witnesses, we are entirely satisfied that the respondent has discharged the burden that the first appellant employed dishonesty to secure the TOEIC ‘speaking and writing’ score that he relied upon in support of his application for entry leave to enter the UK as a Tier 4 student that was granted and was valid until 27 June 2013.
137. It follows that we dismiss the appeals.

Notice of Decision

138. The appellants appeals are dismissed.

V. Mandalia
Upper Tribunal Judge Mandalia

Judge of the Upper Tribunal
Immigration and Asylum Chamber

7 March 2024